Proceedings of the Conference on
^^ NINE-DOT BRAILLE"
June 19, 1964
AMERICAN FOUNDATION for the BLIND
15West 16th street, NewYork.N.Y. 10011
REPORT OF THE PROCEEDINGS OF THE CONF''RENCE
ON "NINE-DOT BRAILLE" - JUNE 19. 196^
HELEN KELLER ROOM
AMERICAN FOUNDATION FOR THE BLIND, INC.
NEW YORK CITY
Dr. S. C. Ashcroft, Chairman
Department of Special Education
George Peabody College of Teachers
Mr. M. Robert Burnett
American Founiation for th*^ Blind
New York, N.Y.
Dr. Thomas A. Benham
Professor of Physics
Editor, Sci'^nce for the Blind
F;r . John Covici
Assistant in Developing "Nine-Dot-
■\'ay n e University
Dr. Emerson Foulke
Deportment of Psychology and
University of Louisville
Mr. Kenneth R. Ingham
Massachusetts Institute of Technology
Mr. Howard M. Liechty
Managinp- Editor of the "New
Outlook for the Blind" and the
"Matilda Zi egler Magazin<*"
Member of f^e Board of Trustees
American Foundation for the Blind
New York, N.Y.
Dr. Abraham Nemeth
Assistant Professor of Mathenat-'cs
Unive^^sity of Detroit
Dr. Carson Y. Nolan, Director
De-oartment of Educational Rese-.rch
American Printing House for the Blind
Mr. Harold G. Rob'^rts, Director
Program Planning Denartrnent
American Found iti on for the Blind
New Yo^k, New York
Mr. Carl T. Rodgers
Pro-ram Srecial ist in Braille and
Other Tactual Education M.^.tcrials
American Foundation for th? Blind
New York, N.Y.
Dr. Robert A. -cott
Socia"! Science aid '/fork for the Blind
Russell Sage Foundation and The New
York Association for the Blind
Mr. Robert Strom
Inventor and Irotagonist of
"Nine -Dot -Braille "-Under graduate
Most people are aware, at least in a general way, that there
exists a system of raised-dot writing which is read by the fingers
of persons who do not see well enough to perceive the shapes of the
letters of the Roman alphabet, and that this system was devised
more than a century ago by Louis Braille, whose name it bears. Too
often, the emotionality evoked by blindness obscures the most
fascinating characteristic of Braille's invention. As the following
illustration shows, it was designed and developed with mathematical
precision and symmetry, for the inventor was an accomplished blind
mathematician and teacher.
The June 196^ Conference on "Nine-Dot Braille" was a logical
continuation of the century-long efforts of Braille and his succes-
sors to put the printed word, and all other written terminologies,
within the reach and comprehension of those members of our literate
society who are blind.
It is also a logical parallel that a number of those who
participated at the Conference should be accomplished blind mathema-
ticians and teachers in various institutions of higher learning, for
the braille system has always been to a large extent a function of
education. It is probable that modern science and technology made
.these participants more keenly aware of the magnitude of touch reading
problems than were Braille and his immediate successors.
The complexities of the task before the Conference are perhaps
best illustrated by the second item on the agenda, entitled "Reada-
bility of Nine-Dot-Cell Characters; Type Scale; Tactually Discriminable
Space Difference Between Modifying Dots and Base Sign."
It was not the purpose of the Conference to seek specific
answers to any of the problems of tactual legibility, but rather to
identify all the problem areas concerning the expansion of Braille's
six dot base cell into a nine-dot base cell, and reduce these and
other tactile problems into researchable areas.
Mr. Strom's resume, which includes the reasons why he became
interested in designing a "Nine-Dot Braille" code, and the agenda
topic entitled "Need for 'Nine-Dot Braille'; Orthographic Efficiency
and Space Saving," constitute an ample explanation as to why the
American Foundation for the Blind considered the subject of
expanded braille codes so important as to have called the Con-
ference presently being recorded, as well as a previous one which
was also held at the Foundation on July 6, I962.
The Conference proceedings have been arranged in topical
rather than chronological order to give maximum relevancy to
each of the participant's comments and recommendations, and in
keeping with the topics listed on the agenda which have not, there-
fore, been repeated in this report. Nothing, however, has been
changed, added, or deleted from the participant's statements.
A Working Paper (see Appendix A) was prepared and sent to
the participants in advance of the Conference. It was intended
merely to stimulate discussion. Subsequent to the Conference,
Dr. Nemeth offered to supplement his remarks on the determination
of the intrinsic ambiguity of six-dot and "Nine-Dot Braille"
(see Appendix C),
The soundness of a philosophy of the presentation of
written terminologies in tactile reading form, the logic and
precision of the physical development of punctographic codes, the
strict adherence to information theory in the assignment of mean-
ings to the various characters of a code — these and other aspects
of code development are highly important; but code optimization
will not be feasible until the necessary data on the tactile human
sense have been gathered and organized by research specialists.
It was the hope of the Conference participants that
extensive scientific investigations on touch reading would grow
out of the Conference discussions. In the interval since the
Conference, at least one major research study has been started.
Dr. Emerson Foulke, of the University of Louisville, a participant
in the Conference, has received a grant from the U.S. Office of
Education for a project entitled, "The Development of an Expanded
Punctographic Code for Tactual Communication."
The study, according to Dr. Foulke, is not being undertaken
with the thought of replacing the present braille code, but to
determine the advantages and disadvantages of an expanded code.
Many of the Conference participants will serve in advisory capaci-
ties during the project.
The American Foundation for the Blind, which provided some
financing of the project, is pleased that one research project
is already underway and hopes that still others mav be undertaken.
The Foundation wishes to express deepest appreciation to all
of the participants for the diligence and vigor with which they
addressed themselves toward fulfilling the purpose of the June 19,
196A- "Nine-Dot Braille" Conference,
M. Robert Barnett
TABLE OF CONTENTS
List of Participants
Introduction ' Pages i-iii
Opening Remarks (by Harold G. Roberts) Page 1
Welcoming Remarks (by M. Robert Barnett
Executive Director) Pap-es 1-2
Resume on "Nine-Dot Braille" Code Develop-
ment (by Robert Strom) Page 3
Readability of Mr. Strom's First Nine-Dot Proposal. Page 9
Readability of Nine-Dot Cell Characters, Type Scale;
Tactually Discriminable Space Difference Between
Modifying Dots and Base Sign. Page 11
Need for "Nine-Dot Braille": Orthographic Efficiency
and Space Saving. Page 3^
Questions on Braille Code Design: Philosophy of Pres-
entation of Information to the Braille Reader. Page 50
Participants' Comments on the Duplication of
"Nine-Dot Braille". P^g® ^^
Consensus of the Participants' on Structuring and
Organizing an Effective Research Project on Six and
"Nine-Dot Braille"; remarks concerning Funding. Page 65
Consensus of the Participants' on the question of
Publicity. Pag® S8
"WORKING PAPER" - "Nine-Dot Braille" Conference
Submitted by: Carl T. Rodgers
Specialist in Education
(Braille & Tactual Aids)
Report on "Nine-Dot Braille" Conference held at The American
Foundation for the Blind, July 6, 1962
"The Mathematics of Six-Dot Braille and Expanded Punctographic
Codes for Touch Reading and Writing".
Submitted by: Dr. Abraham Nemeth
Assistant Professor of Mathematics
Mr. Roberts: I am Hank Roberts. I am Director of the Foum^^tion ' s
Program '^'lanning Der.-rtment and I am f^oinp to be here with you mort
of the ^ay. I hpvp no technical knov/ledp-e of the sub.-ject th^t v;ill
be discussed here today, but I am extremely interested and, more
than that, I'm very s^rateful, as all of us he-° are, for qll of you
coming here and r^t"''-"'-C"' ri;dting in this discussion that, I am suT-e,
will prove to be very helrful to many, many i^eople.
Mr. Parrett : Everv time v/e have any kind of a meeting in the
foimdation, about all I do around here is come in and welcome grou-ps
of people. I have such a stock speech that th'- staff has h^-ard i^-
so much that T do sometimes try to change it, but it usually go-s
along the line that it i" not up to me to w.=>lcome you -^ n the se^-se
of helping the Foundation as such. It is, rather for us to thank
you for using our facilities from time to time to try to advance
knowledge about s^^rvices to blind persons or to advance or improve
This particular subject is one that rennires considerable
genius, and I would say, brilliance. Evidently in this room there
is probably almost all the knowledge there is in the United States
of this -narticular subject. Apropos of braille, I remember a little
verse that was once re-irinted in the rUTLOGK ard it w^nt something
Down in braille alley, down on my knees
Tryinp; to find my stylus
Lord help me, please.
Can't you just get a vision of that little boy crawling around look-
inp- for the stilus.
I am also glad that thir plan, this study, has been going on
for some little while now, thanks to Bob Strom's initiative'' , and the
extreme interest that was immediately shown by Miss Gruber and Carl
Rodgers and John Dupress, The reason I am mentioning those two
names in particulc-r is that some of you may be aware that Miss Gruber
and Mr, Dupress are no longer with the Foundation.
Peyord this, the only thing I can assure you as a c-roup, is that
the Foundation is prepnred to pick up whatever your recommendations
are. I notice on the agenda later in the dav, the obvious questions
of where do we go from here?
I'm strictly an amateur, but have had enough coaching from the
Research Staff to Vnow that if research is to be done ripht, it can
be quite expensive and can take a lot of time and involve a lot of
headaches and blood, sweat and tears. Insofar as T can help arrange
the financing and the facilities for whatever the next step the group
recoHuaends should be taken, you can certainly count on me and the
Board of Trustees,
Mr« Roberts; .......Carl has worked very hard in planning this
meeting for today and has really done an excellent job. I am sure
you would all agree that it is with a great deal of pleasure that
I turn the meeting over to you Carl.
Chairman Rodgers : Thank you very much for your very kind words,
Before I introduce Robert Strom, who is going to give us his
resume on his "Nine-Dot Braille" efforts, I would like to say just
one thing if I may, pertaining to myself.
I have attempted to prepare a Working Paper on Research Areas.
Know ye all, I am not a research man, and I do not labor under any
hallucinations of being a research man or being acquainted with
semantics, the jargon of research people, which simply means that
as we go along with this discussion, if some other categories other
than those which I have set down should be included or be taken in
a different order, I certainly will be glad to do whatever seems
most feasible to the group from the standpoint of a research discussion.
I would like now to introduce Mr. Robert Strom, who will sum-
marlBe his efforts on "Nine-Dot Braille".
RESUME ON "NINE-DOT BRAILLE" CODE DEVELOFMENT (by Robert Strom )
Mr. Strom : Thank you, Mr. Rodgers.
I want to give you just a brief historical background as to how
this system originally got started and what eventually brought me
before the American Foundation for the Blind with a proposal of a
"Nine-Dot Braille" system.
It all started in June of 196l, when Mr. Donald Barr, who was
then director of Columbia University's "Science Honors Program",
which was a classroom and laboratory pro2:ram for pre-college students
with unusual scientific aptitude, requested my assistance for a
I was a computer programmer, and I was requested to give pri-
vate instruction to Mr. John Covici, who at that time was a student
in the Mathematics and Computers Section of the Science Honors
Program. I was supposed to ascertain all the problems involved in
a blind person's learning, writing, and applying the techniques of
computer programmers so that my instructions could serve as a sup-
plement to the standard lecture and laboratory prop;ram used in teach-
ing John's classmates in the Mathematics and Computers Course. This
lecture sequence was, for the most part, rather visually oriented.
For me and for the people at the Columbia Science Honors Program,
this was an introduction to the problems and limitations of a blind
person's learning a skill like computer programs. It was from this
encounter with "trying to teach a program that involved writing of
technical symbols that the idea of improving the braille system,
which eventually led to the "Nine-Dot Braille" proposal, was conceived.
Then in the course of instructing John, I familiarized myself with
braille, with Standard English Braille, Grade Two and with the 195^
edition of Dr. Nemeth's Code for Braille Mathematics. I examined these
codes carefully, trying to ascertain the logic behind the choices for
the configurations. Many seemed at first sight strange.
I had the conception that any system of communication, whether
it was verbal or written, should contain a certain amount of redun-
dancy of information; that is, whenever two signs of distinct sig-
nificance occur, they would be less likely to be confused if there is
more than just one perceptual difference between the signs like, for
example, in spoken English, the words "bat" and "pat", where only the
voicing of the consonants serves as a cue for distinction, are more
easily confused than "bat" and "cut", where there is a difference not
only in consonant quality but also in the vowel.
Yet, in examining braille, there were many situations where the
only other graphic separation between meanings was one of position
with respect to blank cells such as the sign for "were" and the sign
for the left parenthesis; that is dots 2, 3, 5, and 6, where the dis-
tinction of meaning is based upon position of the sign for "was" and
the sign for "by" where the distinction is made on the basis of wheth-
er there is a space after the contraction. Or there were some char-
acters where the only difference in perception between them was their
vertical position within the cell. These seemed to be more percep-
tually confusing than situations where there were more distinctions
between the contractions. However, the need for redundancy in this
perceptual legibility problem and the need for space saving and time-
saving contractions which speeded up reading often ran in opposition
to each other and had to be balanced off.
When there are only 64 possible character configurations, there
is a bad problem resulting from the fact that you do not have very
many configurations to choose from. The result is that in literary
braille, where we have a large amount of cueing from the context,
what happens is that the orthographic redundancy is sacrificed for
the sake of space saving contractions and, hence, we have a large
number of situations where the same sign is used to mean different
letters or characters, depending upon surrounding conditions.
In mathematical braille there is less cueing from the context.
We, therefore, had to make more different contractions, and we ran in-
to problems where there are multiple cell configurations five or six
cells long to represent one symbol in mathematics. Also, one of the
patterns that ran through both kinds of braille codes was that of
three indicator dots followed by a six-dot base cell basically a nine-
dot configuration in two cells.
So this was my first examination of the braille system that was
used at the time I was teaching John. There were many problems with
orthography as far as representing a lot of the information in the
computer course and in the course in matrix alegbra, which is also
being taught. These were mostly problems of orthographic inconvenience,
I decided to experi ion of the braille system,
that is, a way of expanc...^^ „ ..,.._ ., „ possible one-cell config-
urations by a factor of eight. I felt that it was possible to both in-
crease the number of possibilities for space saving contractions and
also increase the number of different configurations and, hence, elimi-
nate the overlap where we have different meanings assigned to these
same configurations of dots.
What I did at that time was the following: I retained the sixty-
four basic single characters of Grade Two braille and I also retained
most of the formerly developed concepts of how to make contractions:
whole-word contractions, part-word contractions, single-letter whole-
word contractions, initial-letter and final-letter contractions, short-
form words, and then I added dots 7» 8 and 9 to the right of the base
cell. I transformed the contractions which consisted of dots 4, 5»
6, plus a ceil, to a nine-dot contraction containing the same base
cell plus the indicators in position 7, ^ and 9 of the same cell. For
example, "ation" which was dots 6, l-3-^-5i became dots 1-3-^-5-9.
On the basis of this and also on the basis of an elimination of
all the situations where the same sign stands for two different config-
urations, I drafted a first tentative copy of both a literary and a
mathematical braille system, more or less on my naive intuition at
that time, and I then searched for organizational aid in supporting
the construction of slates so that I could produce some "Nine-Dot
Braille" in sufficient quantity for testing it.
There were many nroblems in looking for help before I came to
the American Foundation. Part of the problem was a fear of a new
system of braille that would just add to the long history of changes
that have been made in braille and part to the prevalent conception
that psychological experiments had proven that six dots were the largest
number of dots that could fit into a single cell.
As to the second problem, I consulted Professor of Psychology
Dr. Jerome Brunner of Harvard, and he explained that the experiment re-
garding threshold of immediate perception and the maximum number of
single perceptual elements that could be perceived in a single glance
did not apply to the limitation of the size of the braille cell. It
merely meant that when subjects were presented with patterns of dots
on a tachistoscope in the case of visual images or with an instantane-
ous glance and tactual images, there was a certain limit to the sub-
■iectr. ability to identify them as being composed of a certain number of
dots. But this did not mean that a person could still not identify
patterns composed of more dots. He illustrated with the example of
Chinese, where some of the basic signs consist of more than nineteen
strokes; yet people can read Chinese very quickly even though they may
not be able to tell at a glance how many strokes are in the symbol that
After some effort at locating help, I eventually came to the
American Foundation and the very dedicated help of Mr. Rodgers, and he
arranged a conference of specialists in July 1962 at which some of you
who are here now attended. You can recall the recommendations that
were made at that time, which included the following:
1. "Nine-Dot Braille" has potential for increasing the
scope and flexibility of tactual reading. It may be
said that it could become a successful general-purpose
reading system. We must find out whether it can, by
experimenting net with the present "Nine-Dot Braille"
that was offered before the last Conference, but, rather,
with a group of systems which share the "Nine-Dot Braille"
concept. In short, we should not stick to that one
tentative proposal of assignment of meanings that was
presented at that particular conference,
2. "Nine-Dot Braille" has potential weaknesses which may
aggravate the problems of tactual legibility. For
example, bunched contractions should be avoided. How-
ever, it was not known whether the appearance of illegi-
bility of these contractions was the fault of the con-
tractions themselves, i.e., the dot configuration;
whether they were the fault of the configurations com-
bined with their meanings, v/hether it was the fault of
the unevenness of the slates that had been produced in a
hurry for that conference; or whether it was the fault
of the spacing value, which was .080" used in that version
of braille. There were many different reasons, and it
was not necessarily the "Nine-Dot Braille" concept which
was at fault. We did not know which.
3. "Nine-Dot Braille", whenever feasible, should continue
the system and the pattern of Grade Two braille as much
as possible so that people who are currently reading
Grade Two braille do not have to make a violent tran-
sition to "Nine-Dot Braille" if the system is ever re-
leased. In particular, what was implied was that the
"Nine-Dot Braille" cell should be considered as composed
of tv;o parts: a 3 x 2 submatrix of the cell, which is the
basic six-dot cell, plus three indicator dots which would
serve as a modifier to the six-dot sign that is inside
the nine-dot cell. It was also implied that the pattern
of initial-letter and final-letter contractions should
4. Indicator dots should be placed at the left and not at
the right of the base-sign part of the cell, especially
in mathematics where we had situations, when it was at
the right, that a sub-2 looked like a 2-sub when read
left to right.
To experiment with the spacing values, the Foundation created
three different slates which we have here. The spacing values of
these slates are .O85", .090", and .095" between dots in a cell.
Revisions to the "Nine-Dot Braille" system have been made since the
time of the last conference. Sample material of literary amd math-
ematical braille codes were made on each of the slates. We have
those. And the results were shown to Mr. Rodgers prior to this con-
ference, and he imrr.ediately pointed out, after careful examination
of the "Nine-Dot Braille"key and of the samples of material, that we
had a problem in legibility again which may be a result of the pro-
blem in spacing.
Basically, it is the following: what would seem to be a logical
division of the "Nine-Dot Braille" cell into modifier and base sign
based upon the dot numbers may not always be psychologically sep-
arated into modifier and base sign according to the pattern that we
have created. For example, remember that dots 7,o,9 are now on the
left and dots 1,2,3,^,5,6 are now on the right. We had a contraction,
dots l-2-3-if-7-8 (dots 7-8 "p"). Thpt would seem at first glance
to be a good initial contraction based upon a "p" namely dot 7-8 p.
However, when you actually read the contraction under your own finger,
it turns out to look like an "er" sign with a dot trailing off at
the right, and this would be a confusing perceptual cue because "er"
does not have any phonetic resemblance to "p".
Of course, one might argue that you could still use it to re-
present, for example, a "per" sign. However, this would violate the
pattern that we have set up, namely, that initial-letter contractions
should always have the cue based upon the initial letter of the word,
that is the "p". However, if we want to build a new contraction where
this particular configuration is based upon the "er" sign, we would be
violating the pattern that the indicator dot is on the left and not
on the right. So we have a dilemma here.
The question that was raised is basically whether there is a
natural psychological breakdown of a given cell into the parts, a six-
dot and a three-dot subcell. Or is it just that the particular appear-
ance to Mr. Rodgers may not be the same as the particular appearance
of the cell to another blind person reading the system? We have to
determine this fact.
The second problem is what to do about this. Should we relax
the constraint about keeping within th« same pattern precisely of
Grade Two braille or should we modify the spacing values within the
cell so that the "Nine-Dot Braille" is no longer a square cell but
instead a cell where there is more spacing between the three dots
at the left than between the six dots at the right in order to sep-
arate the characters more easily.
The problem is what kind of alternative solution with regard
to overcoming the problem does this pose and how does this affect
the future of "Nine-Dot Braille"? Basically this, combined with the
problems of legibility, are among the things that I would like re-
commendations from this Conference today.
READABILITY OF MR. STRCM'S FIRST "NINE-DOT" PROPOSAL
Mr. Ingha m; Could I make a couple of points, having been at the last
Conference? The first thing I would like to point out is that a
resume of the suggestions of the last meeting as read here, I think,
conveys a wrong impression. That is that these suggestions were by
no means that firm. I think the only firm statements made were that
"Nine-Dot Braille" could be read as it stood then, and that it should
be researched as quickly as possible by a large number of people to
see what could be done. I mean, we could not really tell at that
time, in a one-day meeting, what these values and problems were; whe-
ther, for example, dots 7,8,9, should be placed on the left or right.
This was not at all firm. I think this is slightly different. I
wonder if Tom Benham has similar feelings.
Dr. Benham : As I remember it, we didn't come up with any specific
recommendations about the aspect of it. It seemed to me as though
we were talking about the possibility of getting the material in the
hands of a lot of people so that it could get some widespread eval-
uation instead of just by us.
Mr. Ingham ; The only unified impression we all had is that, as it
stood then, we could definitely read it.
Chairman Rodgers ; Mr. Liechty, as long as we seem to be reminiscing,
would you care to offer any comments?
Mr. Liechty : It's my impression that there were no firm recommend-
ations. I think I agree with these two gentlemen who have said that
it appeared very legible. I was, I recall, amazed at how quickly you
people, who read by touch, picked it up and followed Bob Strom's
reading and prompted him at times when he hesitated a little bit. So
it seemed quite legible as I observed the reading; also that it should
definitely be tested by many more people.
Chairman Rodgers ; Now I would like to give a little sequel to that,
particularly with respect to the testing by many people.
It was, as I recall, the consensus toward the end of the con-
ference that "Nine-Dot Braille" should not constitute a new system, at
least insofar as literary braille was concerned. The point was made,
I believe by Mr. George Meyers, that during the last forty years or
so we extended the situation to include Moon type and other forms of
raised lines; that blind people have been subjected so far to many,
many changes which, from a surely pragmatic point of view, would in-
volve a lot of interruption in the progress that has already been made
with respect to the development of touch reading and that, therefore,
at least insofar as the literary braille is concerned, it should
remain pretty much as it is now, with the possible exception of
putting the modifying dots at the left and as part of the nine-dot
cell. Thus the contraction for "mother" would not read "m" plus a
middle dot but, rather, would read "m" preceded by a closeup dot
which formerly, in a two-cell system, would have been dot 5
Mr. Inp^ham; Just a couple of more things. Bob, did your material
at the last conference have the modifying dots? For example, in
the case of "mother", was dot 5, which is traditional, following
the sign then?
Mr. Strom ; It was at the right originally.
Mr. Ingham ; I think we rebelled against the idea.
Chairman Rodgers ; Yes, they did definitely.
Mr. Ingham; As far as the random suggestion we made about the
nine-dot system replacing the literary braille, you are right there.
But as far as any other kind of braille, such as math braille, we
felt that there were no holds barred and really we couldn't tell
whether the dot should be on the right or the left. Or, indeed,
whether the present way of writing was at all an efficient way in
Chairman Rodgers; Right; and, indeed, that would follow along the
traditional development of braille anyway, because some of the time
in mathematical braille the modifying dots are at the left and some-
times there is a modifying dot, or indicator, at the right of the
READABILITY OF N INE-DOT-CELL CHARACTERS. TYPE SCA LE: TACTUALLY PIS-
criminable "sface difference between modify i ng dots and base sign .
Cha irman Rodders ; The first thing I see on the agenda is "Readability
of "Nine-Dot Cell Characters," and under that we have "Type ocale,
Tactually Discriminable Space Difference of Modifying Dots, that is,
between modifying dots and the base cell.
Dr. Nemeth, would you care to start the discussion?
Dr Ne meth ; For the sake of uniformity, I hope that we can get a
standard way of talking about dimensions of braille, because trying
to read different ideas that people have is very confusing.
There are essentially three dimensions. There is inter-dot,
inter-cell, inter-line. Inter-dot means the spacing between the
center of one dot and the center of an adjacent dot in the same cell.
By inter-cell, I mean the spacing between the dot in one cell, and
the corresponding dot in the other cell. Grade Two, this is about
250 of an inch. In various things that I have been reading, people
have sometimes been talking something like I6O thousandths, and then
1 understand the spacing between the far side of one cell and the near
side of the next cell. This is very confusing. We should stick to
Inter-cell spacing is 250 thousandths of an inch standard, that's
all. Inter-dot spacing is 90 thousandths ofan inch, that's all.
Inter-line spacing is ^00 thousandths of an inch.
Covici : That's corresponding dots.
Dr. Nemeth : That's right. You should always use this nomenclature..
Chairman Rodgers: To satisfy Dr. Nemeth and anybody else who may
disagree with that nomenclature in the Working Paper, I used both.
- In some instances, it may be necessary to use both for clarification.
I personally prefer the corresponding dot system — distance between
corresponding dots of adjacent cells and adjacent lines.
Dr. Ashcroft : It isn't quite true that it is from the far side of one
to the near side of the other. It's from center to center m the case
Dr. Nemeth: That's true.
Chairman Rodders; Yes, it's from center to center in the case of
the dot .
Dr. Nemeth; There's a fudge factor there, I agree.
Dr. Ashcroft ; There's a good reason for that, I think, because of
the base diameter of the dot.
Dr. Nemeth ; There is also an intra-dot dimension that I did not
^n?!^°^n ! f"*! ^°lff dimensions of braille; intra-dot, inter-dot,
inter-cell and inter-line. Intra-dot means the base diameter and
tM« if . .*^^?^ ^°" ^^^ everything else can be inferred from
this. There is no clarity gained by reverting to another system, because
people are mentally oriented to one kind of set of dimensions, aid
r?MnV^^>,^r^^ °*^^'' measurement from a standard set of dimensions, and
I think that If you want to have universal understanding, you must stick
to one set of specifications. That's the first thing.
The second point that I wish to make is this. When you prepare
experimental material, I think it would be lethal to try to put all
i°: i^^V^'*^^^^^^ in at once. I notice that in the material of which there
is only one copy, which you people studied at the last session, you
have regular standard English braille where the inter-dot spacing is
90 thousandths and then you have a nine-dot braille where the inter-
dot spacing is 80 thousandths. I don't know whether there's also a
discrepancy between inter-cell and inter-line spacing.
Besides that, the standard braille has I89 contractions and short-
form words whereas "Nine-Dot Braille" has lots of other contractions.
So this makes the comparison very difficult. No one knows what they
are really comparing. "^
Chairman Rodgers: May I say a word as to that. In the first place,
the first conference was not structured as a well organized research
fu°^®^u J\*^® second, there was no nine-dot writing equipment other
I' m! ! *^^* ^^^ ^*^°°^ ""^^ successful in having constructed for
him. That meant
Mr. Covici ; That's not quite true.
Dr. Nemeth; That's not true. I could have sent you a nine-dot slate
if you wanted it.
Mr. Covici ; (She New York Point slate is 3x3*
•Editor's Comment: The New York Point System is two dots high. Theo-
retically there is no limit to the possible horizontal extension of New
York Point characters; however, in practice no character extended horizontal"
beyond four dots.
Dr. Nemeth : Yes
Chairman Rodgers ; That would do it, but it wasn't available to us.
Mr. Ingham : That's minor point anyway.
Chairman Rodgers : Then you have an assignment of going around and
seeing if you can collect one,
Dr, Nemeth : My question is why not?
Chairman Rodgers : Your question has been noted, sirl There was no
slate available period; other than the one which Mr. Strom had, which
would have meant that he would have had to write the whole thing, the
sample material and everything else; he would have had to peck it
out on the one slate, and in order to save time so that the conference
could take place at the assigned date, the standard braille, the
regular braille material was written on a braillewriter and the pur-
pose of the other material was simi'ly to get a consensus as to whether
it might be feasible to carry on from there with respect to research,
and so on. We are not going to decide anything here at this meeting,
gentlemen, as to which spacing variables and which set of comparisons
is going to be the best. That comes at the end of the agenda, where
you're going to have to think of setting up a well structured research
That is the reason for that. We agree with everything that Dr,
Nemeth has to say and we hope that, as the research project gets
underway, all of these universal norms will be adhered to and that a
nomenclature will definitely be agreed upon. As it is now, there is
no specific nomenclature agreed upon because if you read some records
of investigations — for example, those of the Commission on Uniform
Type — they give you those that Dr. Nemeth has proposed. But if you
read other experiments and other records, it is a little different.
Dr. Nemeth : Let me make another point, the system of modifiers and
full cell bodies — I don't know what else to call them — modifiers
and cell bodies, is a very good system when it works for saving space.
We have, from Mr. Rodgers' spacing values, the fact that a 40-cell
line of standard braille will be reduced to approximately a 32 — or
a 33 — cell line in "Nine-Dot Braille", Am I right about that, Carl?
Chairman Rodgers : It varies. I would prefer to use the cell totals
per writing area. I think it was reduced, depending on the size of
the type, from 1,000 to 725.
Mr. Covici ; Incidentally, that was the lowest.
Chairman Rodders ; Yes. 725.
Dr. Nemeth ; Yes, but that is if you have high density braille, you
know, with small inter-dot spacing,
Mr. Covici : No. That was the worst.
Dr. Nemeth ; I see.
Chairman Rodgers ; That was .090" or to use your terminology of
corresponding dots, .3^0".
Dr. Nemeth ; I see. Anyway, the point is this, there Is approx-
imately a 20-to-25 per cent loss of numbers of cell per line or per
page or per area, or what have you, in that neighborhood, in that
order of magnitude. Let me confine myself to a set of contractions
in the order of magnitude of I89, like in Standard English Braille.
Let's keep that variable constant. Then this means that your nine-
dot system will have to be overall — ■ not in the specific case, you
understand, but overall -- approximately 33 per cent more efficient
just to keep the space you already have without any gain on it. This
means on the average that you will have to have approximately eight
to nine dot-cell combinations per line just to break even with
ordinary braille, because if you were to write ordinary text or any-
thing like Bob had on the previous one, such as circuit statements or
ordinary text where the contractions come at random, you don't get
anywhere near the number of eight to nine cell-forms per line. You
see, if you have 32 cells and you want to get up to ^0, you need those.
If you have this system, it's very fine to have these long lists
of contractions, but unless you get eight per line, you're not going
to break even in space with your old-fashioned six-cell braille. This
is the most important point to consider, I believe, even superseding
space values and what have you, because if you're not going to break
even with ordinary braille as far as spacing is concerned, you're
not going to have any advantage. I wish you people would kick this
around a little bit.
Mr. Ingham : I would like to, with respect to English braille.
Dr. Nemeth ; All right, I would like to comment about mathematical
braille. If you use the modifier-cell-body combination, you have at
your disposal eight modifiers including the "empty" modifier which is
no modifier at all. As you know, Ken, from being on the math com-
mittee with me, we have what, some 30-odd braille indicators?
Mr, Ingham ; At least.
Dr. Nemeth ; Some 30-odd indicators and we need them all. And the
nine-dot cell is not going to help you. It is going to help you im
this respect: First of all, one of the things we have in our math
code is the possibility of performing five alphabets and within each
alphabet we have the possibility of designing ordinary type, script
type, bold face type, or italic type, which means for each letter
you have 20 possibilities which is far outside the limits of your
"Nine-Dot Braille" modifier product combination, you see. So the
trouble is they are just as bad" in mathematical braille as they are
in regular braille.
Mr. Ingham ; I Just want to point out that there is an over-simplified
way of looking at it. There are certain advantages which the "Nine-
Dot Braille" might have in mathematics.
Chairman Rodgers ; Do you want to get those on the record, please?
Mr. Ingham ; I can't think of them all offhand,
Dr, Nemeth; You would have to have eight modifiers per line to make
it advantageous, remember,
Mr. Strom ; We have at least that.
Dr. Nemeth ; In mathematics? No, you don't.
Chairman Rodgers ; Would you care to delineate what those are?
Remember that we're relating this right now to spacing of type scales;
in other words, physical compactness and so on. Are you ready, Robert?
Mr, Strom ; One point I wanted to make with respect to mathematics.
Chairman Rodgers ; Because if you're not, I think somebody called me.
Yes, Dr. Foulke. Go right ahead.
Dr. Foulke ; I was a little bit surprised, I guess, to hear so much
of the conversation related to literary braille. If I understand it
correctly, the whole discussion of "Nine-Dot Braille" came up in
connection with the need for a more elaborate symbology to express
.ther languages, didn't it? Mathematics, computer?
Chairman Rodgers ; Both,
Dr. Foulke ; You see, this problem did not arise in any effort to
solve any problem relating to literary braille.
Chairman Rodgers ; Yes. Excuse me please. Let me tell you a little
bit of the background.
At the first conference, Dr. Foulke, it was emphasized that
since the literary braille readers would probably rebel at any new
system or at what would seem to them psychologically a more com-
plicated system because of nine dots instead of six, it was decided
that if there was going to be a "Nine-Dot Braille" system, the stage
of development would begin at the technical levels; in other words,
mathematics, music — the special code levels, shall we say. But
at the same time the possibility of extending "Nine-Dot Braille" to
the ordinary literary code was not excluded at all.
That is point number one. Point number two. When Mr. Strom
presented the revised "Nine-Dot Braille" code that he had prepared it
seemed, at least to me, that because of there being no space differ-
ence between modifying dots and the regular braille cell dots, what
resulted was an entirely new configuration in a good many instances,
and you will see that later on. Mr. Strom and I decided — right,
Mr. Strom, you can correct me on this -- that the same blending of
modifying dot with regular base characters would also result in
mathematical braille and, therefore, while we do believe that the
developmental stage, so far as the structuring of a system is concerned,
would begin at the special code levels, for practical purposes we
took his literary key as an example because the type scale problems
are going to be the same both for mathematical and literary braille.
So we just took that one as an example. Do I make it a little
Dr. Foulke ; A little clearer except that I am not really sure that
the spacing problems are the same. For instance, it may be that space
economy is a more important factor in literary braille than in other
codes. Space economy is, of course, important, I'm sure, in all of
the codes, but it may be relatively more important in literary braille.
This means that in the case of literary braille certain other arrange-
ments might have to be made.
Mr. Ingham : Could I make a comment quite relevant to this? This is
the fact that no mathematical text that I know of is strictly math-
ematics. Unless you're prepared to switch back and forth between two
different types of writing, you're just going to have to be able to do
all sorts of literary material in the nine-dot system. So I think
it's a foregone conclusion that you have to consider it as literary
material as well as mathematics.
Chairman Rodgers ; I don't believe you can separate the two.
Dr. Nemeth ; Unless you have one of these keys where they have nothing
but theorem and proof and those are the only English words in there!
Dr, Nolan ; Let me make a more general statement, Mr. Rodgers.
It seems to me that we are all here because there is a problem in
communication; that the present system is not versatile enough to give
us the degree of communication required, and that we want to consider
a modification of the system, namely, going from six to nine dots
as a possible way of increasing the versatility. The line of con-
versation to this point has been to give reasons why this won't work.
We are just on the threshold of the whole business. I think that if
we're interested in exploring increasing versatility of tactual com-
munication, it's not going to do us any good to say it won't work for
this reason^ it won't work for that reason. We haven't even reached
a point where we know whether the characters are legible. This is an
elementary thing without even getting to the point where we use these
symbols symbolically or we use the stimulus patterns symbolically.
So to dwell on these very practical problems, which I admit are of
great importance in the light of the very little information we have
about the potential of "Nine-Dot Braille" - we haven't studied it all -
I think is a little premature.
This is the thing that I confess I often find irritating, because
when people advance new ideas and want to explore, there are always
those around who say, "Well it won't work." This has been the history
of science. There are always those people who sit around and say "No,
that won't work," but people go ahead and explore anyway in spite of
these, and I think that our conference ought to pursue it in the
latter light rather than in the former.
Chairman Rodgers ; Agreed.
Mr. Strom : I agree.
Mr. Covici ; Yes. In the light of that, I think what should be done is
to see what kind of system is readable and do a lot of research on all
the variables and then see whether a book takes less space or whether
it's easier to transcribe or whether it's easier to read, which is
the main point.
Chairman Rodgers ; If we have exhausted the topic of type scale as
such, we can easily go on to another topic. Before we leave type
scale, however, I would prefer to give each participant the opportunity
to express his opinion with respect to the importance of spacing
variables as regards "Nine-Dot Braille", because there is one very im-
portant factor here; How large should the type be in order, on the one
hand, to be meaningful to the reading finger, and on the other, how
small may the type be in order at least to break out even with six-
dot braille, be it literary or mathematical or both, from the stand-
point of physical compactness?
I'm going to start over here at my left and let each one express
his opinion on the specific point of needs for "Nine-Dot Braille"
development with respect to spacing variables.
Mr. Strom ; Yes. The first point is in relation to what you said
about getting "Nine-Dot Braille" to break even with respect to purely
the matter of physical compactness. One thing that has to be agreed
upon is that you cannot specify merely from knowledge of the formal
measurements of your cell how "Nine-Dot Braille" is going to fare
against Grade II braille with respect to comparison. It's always
going to depend upon a large number of factors that are independent.
I wish to mention that we had some discussion with respect to ex-
perimenting with a new kind of breakdown of the cell, i.e., having
two different kinds of inter-dot measurements, that is, not have a
square cell; have one kind of measurement for separation between in-
dicator dots and base-sign dots and another kind of separation be-
tween the individual base-sign dots and consider these as two dif-
ferent spacing variables within the cell.
Chairman Rodgers ; For what purpose?
Mr. Strom ; For the purpose of making it easier for the finger to
discriminate between what is to be the modifier within the cell and
what is to be the base cell.
Chairman Rodgers ; The base character?
Mr. Strom ; Right. And this thing came up as a result of a number of
perceptual confusions which should be noticed on the key that every-
body has in front of him, Mr, Rodgers, do you happen to have an
example of an easily locatable instance of this?
Chairman Rodgers ; If you will give me a little time I will look
Mr, Covici; I have one right here. The one for "her",
Mr. Ingham ; Where is that?
Mr. Covici ; It's on the second line, very near the bottom. It's
the second column.
Mr. Strom : Yes.
Dr. Nemeth ; I have it,
Mr. Strom ; Right. Does everybody have that? Okay. The idea is
that since I had the notion that a lot of personal pronouns had
been contracted into one cell, "her" should also be contracted into
one cell. I chose to try out dots 7-9 followed by a letter "h".
Now the trouble with this is that it is probably going to look to a
lot of you like a "z" with a dot trailing it.
Chairman Rodgers ; Yes. I have here examples of characters like
regular braille characters with a spare. Look for the 7-8 dot "e".
Dr. Nemeth ; Where would that be? Is this in some order?
Chairman Rodgers ; Yes. I would suggest that you look for the equi-
valent first, which is a little to the right, and that stands for
the letters "ee".
Mr. Inghajn ; Where is that?
Dr. Nemeth ; It's "in column 2, the sixth entry down.
Mr. Strom ; That's an "e" with the dots 7-8.
Dr. Nemeth ; 7-8 followed by an "e" yes.
Mr. Strom ; It probably looks like an "f" with something after it.
Dr. Nemeth ; It does not. Let me make this comment, Carl, for once
and for all.
Chairman Rodgers : Wait a minute. Let's examine a few of these. Now
we have a stenotypist here, by the way, and every word that you are
saying is being recorded. I would suggest that we can't get into an
argument of what looks like what to individual "A" and what looks
like what to individual "B" except perhaps to say that "X" number of
individuals will take the 7-8 "e" as an "f" with a spare and "X-Y"
individuals will take it as something else, but certainly you have
no differentiation there between the modifier and the real "e" that
it is intended to be.
Dr. Nemeth ; May I say something?
Chairman Rodgers ; Yes,
Dr. Nemeth : The comment is this: What Carl says is perfectly-
correct provided no one has preliminary instruction. But if you
know that the structure of the cell is modifier plus body, then the
interpretation of the "ee" combination as an "f" with a trailing
dot is not an error, it's a blunder. It just can't be interpreted
Mr. Ingham ; I could make some points here which might be enlightening.
We have at the lab a machine which can braille with a density of
8,000 dots per 10" squared and I was testing this question with this
machine. I talked to Bob Strom a couple of weeks ago about it, and
I suggested this very thing of making one column closer to the
other. Then I went ahead and played around with this machine and
thought a great deal about it and I decided that it would be ex-
tremely harmful to do it. The reason simply is that I could not dis-
criminate when there were several in a row between that separation and,
say, the separation that would occur between two cells, I got highly
confused, and I think we really should make the difference not between
columns but between cells. Make your cells distinctly apart, and I
don't think you would have any problem.
Mr. Covici ; Two points. First, three contractions, as Abe said, at
first glance could look like anything you want. But if you know that
they're supposed to represent a certain symbol, whether the modifying
dot is before or after, or there is supposed to be a blank or whether
it's so and so when you read it, you just have a table type thing
and naturally it comes out. When you're going along reading, you
don't say "Well, this is a modifier and this is a body." Only when
you first lean it, do you think this way, and it may be nice to think
of it that way as a mnemonic aid. Even though the cell used in the
first "Nine-Dot Braille" conference was not so well spaced as the
present one, it was quite legible and people were amazed, and I was
too, when I first saw it. I think that the original cell spacing
might be included in one of the research projects among many other
variables but I'm not sure of its value, I think this "x" decimal
of an inch may be equal to zero with very little loss in legibility
or in anything else.
Chairman Rodgers ; I'm going to turn the floor back to you folks, but
first I would like you to look at some things that I could not find
before and I have them labeled here. Samples of "Nine-Dot Braille"
characters that seem to equal regular braille characters but, of
course, they mean something else. Look at the sign for "ist".
Mr. Covici ; It's on the first page on the right hand column.
Dr. Nemeth ; I have it. It's in the middle of the third column. It's
the fifteenth entry down.
Mr. Covici ; I found it. Yes, indeed.
Mr. Ingham : It looks like an equal sign.
Dr. Nemeth ; No. It's 7-8 followed by the "st" sign. Let me say
this: Bob has been very consistent in making every one of his con-
tractions three dots wide. Am I correct about this, Bob?
Mr. Strom ; So far.
Dr. Nemeth ; I mean I haven't been going down the list, but so far,
Mr. Strom ; The ones that are two dots wide are not contractions but,
rather, regular letters of the alphabet.
Dr. Nemeth ; That's right. So every contraction which is three dots
wide is not subject to a misinterpretation of any kind, of the kind
that you were talking about with the 7-B "m" versus the thing moved
over to the right.
Chairman Rodgers ; The "st" sign.
Mr. Ingham ; What is the matter with the sign?
Chairman Rodgers ; This sign looks pretty much like a "gh" sign
followed by the letter "a" to me.
Mr. Covici ; Only when you first glance at it, but when you look
at it —
Dr. Nemeth ; Modifier-body tells me 7-8 "st".
Mr. Ingham ; Bob, what you should do is give me or 10,000 other
people a couple of weeks to read this. We can't tell. This is the .
worst kind of test.
Chairman Rodgers ; Right. This is not a test; this is a discussion.
Mr, Ingham ; But you can't gain any information. It's just faked.
Chairman Rodgers : This is a discussion and then these things should
go on the record, because while it's true that right now we have
a very erudite group of people here who would be perfectly capable
of indulging in beautiful mental gymnastics in order to read, the
fact is that other people around the Foundation to whom I've shown
these things seem to have gotten the same reaction. So I just want
to bring that out and put it on the record, and with that I turn the
floor back to you gentlemen,
Mr. Ingham ; If you're asking us on our particular level if we can
read this, the answer is yes. If you're asking us in comparison with
these other people, I think you are losing. I think this stuff should
be going to thousands of people in test form.
Chairman Rodgers ; It's a matter for research. I think we are all
in agreement with that; right?
Mr. Covici ; Yes, This is an open question and subject to review.
Dr. Nemeth ; May I make the following suggestion, that when you have
this thing tested, you do not say, "Well, here it is. Can you read it?"
It should be precisded rather by a preliminary informational session
where the potential reader will be told that "When you have a nine-
dot configuration, you should regard the first three dots as modifier" —
Mr. Ingham ; No.
Dr. Nemeth ; Just a minute. Yes.
Mr , Ingham ; Go ahead,
Dr, Nemeith ; Of course. You should regard the first three dots as
modifier and the rest of the thing as the body of the cell. This may
matke a difference. You have all seen the game where you ask people how
to pronounce a word, you spell the word but you stop, you pause between
the wrong group of letters and people pronounce it wrong-^ naturally.
Or you have all seen this gestalt experiment where you have a picture
of a vase. If you look at the vase intently it becomes the figure and
the black silhouette becomes the ground, and so you see a vase. How-
ever, if you are told to look at the blacked-in material as the figure,
then what you see is two silhouette faces looking at each, other and
what used to be the vase is nothing but the ground. What you see
depends on the way you are told to look for something. This is essential,
Mr, InRham : There is experimental work going on now at our lab on
this type of thing. Pictures of different alphabets are being
screened on the oscilloscope off the computer and i^eople are looking
at them. All sorts of things are going on, like letters upside down,
word combinations put in different order, and what has come out
immediately is the fact that by looking at the word as a whole you
can buy a lot, by looking at the cells in sections, you lose. If you
start thinking in terms of modifiers with this conglomeration of
dots after them, then you are going to lose.
Dr. Nemeth : I agree.
Mr. Ingham ; This is quite contrary to what you were saying.
Dr. Nemeth : I agree, but what I'm saying, for example, I tried an
experiment here at the Foundation and it started out as follows:
Someone wrote full braille for the first two lines. This was a
coherent text, you understand. You write regular braille for the
first two lines; then the next two lines was written with leaving
dot three out of every third cell and then the lines thereafter
were just leaving dot three and six out of every single cell. Now,
you know, iust by the context I was able to read this thing down
and there was no dot three or dot six anywhere, but I was able to
read several lines beyond where there was any semblance of braille,
iust from context and interpolation. IVhat you say. Ken, is true.
You should perceive something as a whole, but there comes a time
when you have to make a deduction, and when you have to make a
deduction then you have to know on what basis to make it. You have
to cue yourself to keep going.
Mr. Ingham : Okay. Maybe at first in a training session.
Chairman Rodgers ; I would like to insert a comment of my own with
res"pect to the deductions that we may expect of the ordinary reader
to have to make. After all, Dr. Nemeth, you are not an ordinary
Dr. Nemeth : This was only an experiment. I don't read like that
either, believe me!
Chairman Rodgers : Experimentally or in terms of speed, on the
average, very few people read like Dr. Nemeth and other peor-le
I've known. I could mention them at random. Mr. A, who could read
as f=ist as he could talk. I know Miss Y who could also read as
fast as she could talk, but that is not the average. So it boils dovm to
the fact that you should not be expecting the reader to read through
deductions rather than by cue, the perceptual cues actually on the page
that he might utilize if he so wishes.
Chairman Rodgers ; I've not heard from some of our participants who
have done some very extensive research on tactual discrimination,
on errors in oral reading of braille at the elementary grade levels,
and so on, and I would like to get their reactions, particularly
with respect to Dr. Nemeth's suggestions about how to orient the
subjects to the type and purpose of the test they are to take.
Dr. Nemeth ; They should learn to read in terms of one third plus
two thirds instead of two thirds plus one third.
Chairman Rodgers : That should be explained to them beforehand.
Dr. Nemeth t To be told before, yes.
Chairman Rodgers ; Before the test. I would like to get the reaction
of Dr. Nolan and Dr. Ashcroft and those who have been sitting very
Dr. Nolan ; I will cease to be quiet, on your request. First let me
say I think we are about seven leagues in advance of the question
of cells within dot spacing in the cell. This I think was the ori-
ginal question that was raised and we have gone a little bit afield.
Actually I think there are two problems here. One is having a stim-
ulus configuration with such spacing that people are able to identify
all its elements when they encounter it. This is one problem.
Then there is another problem that has to do with symbolic use
for these different stimulus patterns. With reference to the first,
my guess is that using an increasing number of dots in the braille cell
is going to cause us to make them further apart. This is particularly
going to be the case with younger braille readers.
With reference to the problem of indicating the presence or
absence of a modifying dot, that is, whether a dot 7i 8» or 9 is pre-
sent, there is already a cue. If a cell configuration representing
common usage is present, the space between it and the previous cell is
going to be greater and regular-within-cell spacing will occur when
you have the modifier present.
I wonder about using distance as a cue for the present modifier.
We are working within a very small distance to start with and I
wonder if people are going to be able to discriminate the small dif-
ferences in distance that will be involved.
Chai rman Rodgers : Do you want, Dr. Nolan, to go to the extent of saying
that it is so highly probable that they would not, that it would not be
worth experimenting for the purpose of determining whether "X" decimal
spacing between modifier and base character would be discriminable .
Dr Nolan : I think we are certainly going to have to study this. One
of'the problems, if distance appears to be a feasible cue, is the mini-
mum distance that is discriminable to a majority of braille readers.
This must be defined so that all people can use the system. There are
other cues that one might use. Dot height is a possibility. That
would present some problems for the writing devices that you employ.
You may run into some problems there.
Dr. Nemeth : May I interrupt just for a point of information. How
uniform is dot height from dot to dot in a production run book? Has
anyone measured that?
Dr. Nolan ; It's been measured. I can't give you the information. It's
Dr. Ashcroft ; It's on the order of ten thousandths of an inch, I would
^^, We had great difficulty, when we did the spacing variable study
control, with dot height at 15 thousandths of an inch and we never did
Dr. Benham ; By the time someone has sat on it, why, it's not very
Chairman Rodgers ; Thank you. Dr. Nolan. I think all these comments
are very valuable and we want to keep them on the record. They will
be very helpful in building up a set of recommendations to guide the
research. The reason I stumbled there is because I have no idea what
you folks are going to decide later on as to how to initiate at least
some semblance of a structure for a research project that will come
later. We are a little bit ahead of the game insofar as anticipating
how much the reader should be expected to read by the actual stimuli
which are presented to him and their clarity or lack of it, and how
much by mental deduction or mental gymnastics, whichever you want to
call it. I wonder if Dr. Ashcroft would have a comment on that just
to get it on the record, because invariably that is going to come up
Dr. Ashcroft; Yes. I would like to make several comments.
First of all. I'm going back to the studies of spacing that we
did. As you know, we varied the inter-dot, inter-cell and inter-line
spacing and came up with 27 sets of specifications. It seems to me
that it did not matter what you did very much in spacing; people could
read it and we really had relatively short reading experiences for
them and they accommodated rather quickly. Although Dr. Nolan alluded
two considerations for younger readers and perhaps the necessity of
some larger inter-dot spacing, if I heard him correctly, one of the
things that came out of our dot spacing study was that with younger
readers there was a slight tendency for them to read better with
smaller inter-dot spacing.
Mr. Ingham ; Really? Smaller?
Dr. Ashcroft ; Yes.
Mr. Ingham; That's interesting. They could read better than adults?
Is that what you are saying, with the smaller dots?
Mr. Strom ; No. They could read better with larger dots.
Dr. Ashcroft ; They could read better with smaller dots.
Dr. Nemeth ; By smaller, you mean .085"?
Dr. Ashcroft: Yes. I think it was .085". It might have been as
small as .080".
Chairman Rodgers ; I think the tentative conclusion is that the
inter-dot spacing they read best was .090", as in the case of the adults
Dr. Ashcroft: Yes. It was between-cell spacing. I beg your pardon.
Chairman Rodgers ; In other words, the distance between the end of one
cell and the beginning of the adjoining cell was ,123"?
Dr. Ashcroft; Yes, but I tend to agree with Dr. Nolan that this is
far down the line and it seems to me the first thing to do is to get
some "Nine-Dot Braille" out in this form and have it read widely and
extensively. People can accommodate and learn very rapidly. From
the error studies that I have done, I think that there are many errors
that we don't need to be concerned about. People bring meaning to the
context and the context is a crucial factor in errors and controlled
difficulty level. This is extremely important for children. We can
cause any of us to make reading errors if we make the difficulty
level high. If the difficulty level is appropriate to our reading
ability we can keep errors to a minimum. I just don't think that
this is anything to fear at this point. I would go ahead and get
a lot of material and get it rather quickly. I think it will be
read, and read with facility rather quickly.
Mr. Covici ; Dr. Ashcroft, what was the degree of accommodation
when the inter-dot spacing was reduced to .072"?
Dr. Ashcroft ; There were some statistically significant differences,
but they were small, and again I would emphasize that these people
had a very brief time to accommodate, and I think that they would
accommodate even better if given more time.
Mr. Ingham ; That we could all read the first material at the last
meeting is more than adequate proof of what you are saying.
Mr. Liechty ; Yes. \
Dr. Ashcroft ; People can read anything if they just have the meaning
to bring to it and have a little time to work at it.
Chairman Rodgers ; All the figures on variable spacing which I pre-
sented in the "Working Paper" were only for the purpose of pointing
out the tremendous number of possibilities that there are for vary-
ing these things, and I'm glad Dr. Ashcroft brought up the point
about ability to accommodate to various spacing values, because if
that is the case, with adequate experimentation it would seem to me
that, should nine-dot-cell characters be proved desirable other than
from the spacing values, "Nine-Dot Braille" might be very feasible
from the standpoint of physical compactness.
Mr. Strom ; Mr. Rodgers, I would like to put up a question before the
Conference for tneir opinion.
Chairmam Rodgers ; With respect to type scale?
Mr. Strom ; Right. With respect to the priority in research of in-
vestigations of spacing. The suggestion has been made that it may
be the case that cells will have to be made not square but, instead,
with two different kinds of spacing values within them in order to
provide for necessary cueing relating to modifier versus dot.
Then there is the other opinion — that it may not be necessary
to change the spacing if the cell configurations are suitably en-
dowed with meaning in such a way that the mind learns how quickly to
adjust itself to identify the nine-dot characters as an entire sym-
bol. Also, it may be that in faster reading, modifiers are not even
thought of or that the concept of the modifier and the cell is not
necessarily one that has to be adapted, but there may be some config-
urations which could be pure nine-dot configurations without the con-
cept of modifier.
The question which hence arises is, should we first try to see
how well we can get "Nine-Dot Braille" to be read within the scope
of the spacing values? I would tend to think that, first, we should
see what we can get within the scope of a square cell, adjusting the
meanings approximately and seeing how well they can accommodate to
the cueing the way the square cell is now.
Chairman Rodgers ; Does anybody have any question about what Mr.
Strom said? Is it clear to everybody?
Mr. Ingham ; It's not clear to me in the sense that just before he
made his point I thought the gentlemen, Dr. Ashcroft and the others,
as well as myself, made the point that spacing between different
columns was not only unnecessary but harmful and, therefore, that
question I thought was left. I mean, in what I researched up at MIT,
to use different column spacing is definitely harmful and I under-
stood that the gentleman over here was saying that it didn't mean any-
thing either: The small amount of spacing you could accomplish be-
tween columns would not buy the reader anything.
Chairman Rodgers ; Was it Dr. Nolan?
Dr. Nolan ; I commented on that, I wondered whether the variation in
spacing that was possible would be such that you could get discriminable
differences in spacing. I also wondered if it is really necessary to
actually involve different spacing between modifiers and the others
because, in the presence of no modifiers, the space between cells would
be greater than in the presence of modifiers. This would be your cue,
reduced space to start with.
Dr. Ashcroft; Vacant space.
Dr. Nolan ; That's right, vacant space would be your cue. I'm
getting back to the thing that Mr. Strom asked, is the initial thing
to study discriminability within these patterns? How far must a
dot be from one another in order for people to be able to recognize
all of them present with whatever degree of accuracy is required?
This, it would seem to me, is the initial step to be taken here. Yo
can test this difference of the distances between cells and the use-
fulness of this as an indication of the presence or absence of mod-
ifiers and also the between-dot distances as it affects the ability
of people to be able to recognize all the stimuli present within the
Dr. Ashcroft ; Some of these considerations were studied in con-
nection with the spacing study that was done.
Dr. Nolan : There is some data on this already.
Dr. Ashcroft ; We couldn't vary the inter-cell spacing very fat
because dots ^-5-6 became part of a new cell with the next cell 1-2-
Dr. Nemeth ; That's right. It has a coherence with what follows.
Dr. Foulke; I would like to comment upon something that I think
Mr. Strom either said or intimated. This is that so far we have con-
sidered the possibility of "Nine-Dot Braille" as a kind of logical
extension of the existing braille code, in which case we have been
concerned with modifiers and how they might be used, and we have dis-
cussed the possibility of a rule of consistency regarding such mod-
ifiers which seems entirely reasonable. But there is, at least for
the record, another possibility of considering the patterns that
result from the use of a nine-dot cell as stimulus patterns in their
own right, in which case we might be concerned with the individual's
ability to make absolute identifications of these patterns without
having to go through any kind of analytical process of interpreting
a modifier plus a standard braille configuration. This could be done
and this is another possibility which I think at least ought to re-
ceive some discussion.
Also I would like to make a couple of other comments on some of
the other things that have been talked about. I agree emphatically
with Dr. Nolan that much of what we are talking about is the kind of
questions that can really only be answered by research. I think it
is clear that the research can be done. I think it is obvious to
everybody here that spacing variables — within-cell spacing, between-
line spacing — are all quite important. But beyond listing at
least some of the infinite variety of possibilities that are avail-
able, the only resort is research to answer some of these questions.
It also seems clear that if we are going to answer them, it's ob-
viously going to require the training of some subjects in the use
of codes that we might generate. We can't present them to people
cold and expect them to give us any information about their utility.
Chairman Rodgers ; Dr. Foulke, I think your statement of testing
the reaction to a stimuli per se is an extremely important point with
respect to the development of any orthographic system. I'm speaking
not now as a research individual but as a mere ordinary average
citizen, an ordinary reader. It may be true that we ought to read
two-thirds by stimuli and one-third by deduction or vice-versa.
But I would say that as far as I (and perhaps I speak for the average
reader) am concerned, we want to have as much information available
under our fingers as possible. I don't mean over-cueing or anything
like that, but certainly the orthographic stimuli should be clear
enough that we may avail ourselves of three-thirds of the whole set
of stimuli if we desire, if the information is otherwise not clear
to us. Also, I think we should do a little developing of codes in
a more systematic way for the sake of the nine-dot cell concept per
se and for the sake primarily, of course, of the reader.
Mr. Ingham ; The point I just wanted to make, following Dr. Foulke's
statement, was that I think the way to go about setting up those
codes is to give a couple of people the assignment of writing codes,
keep these questions in mind and submitting these codes, and one or
two or perhaps all of them, should be chosen and distributed. I
don't think these are jobs for a committee to even pass on in terms
of right or wrong.
Dr. Nemeth ; Every good research project starts with someone form-
ulating a hypothesis. This subject is sufficiently complicated for
us to formulate several hypotheses. I wonder if it would be useful
for us to go around and ask people to formulate what their hypothesis
is concerning several subjects in this area: First, the readability
of "Nine-Dot Braille"; second, which space "Nine-Dot Braille" would
occupy versus the space that six-dot braille now occupies; third, a
hypothesis concerning a structure in which each "Nine-Dot Braille"
character has its own meaning versus the structure in which each
"Nine-Dot Braille" character is a two-component symbol of modifier
followed by principal character, I think these are the principal
I think we already have the consensus accepted that the precise
decimal amount of spacing between dots and space, between cells,
and between lines, is easily adjustable too, and that the only im-
portant factor is dot spacing within cells, and I avoid the word
Mr. Ingham ; I think we disagree there. It is the other way around.
Dr. Nemeth : Isn't that what Dr. Nolan said?
Mr. Covici ; No. The other way around.
Dr. Nemeth ; The spacing between cells was not so important as the
spacing within cells.
Mr. Covici ; No.
Dr. Nemeth ; Dr. Nolan, would you edify us?
Dr. Nolan ; I don't think I said any of those things! I said that
the problem was one of legibility of the stimulus pattern and that
this depends on the space between dots in the cells. This is the
principal question. Then you get into the question of what is the
cue to show whether the meaning to be attached to the stimulus pat-
tern is that of the old six-dot cell or the modified six-dot cell.
In this, spacing had been suggested as the cue, one idea being that
the dot 7, 8, and, 9 should be spaced further to the left than dots
1, 2, and 3 are from ^, 5, and 6. And I also suggested that perhaps
the absence of a stimulus, that is the distance between cells, could
serve as effectively as a cue.
Dr. Nemeth ; Yes.
Dr. Nolan ; But the elementary problem is one of legibility of the
Dr. Nemeth ; That's right.
Dr. Nolan; Can you report what you have been exposed to?
Mr. Ingham ; I think this is the point.
Dr. Nolan ; Then you get into the problem of codes. What meaning
do you attribute to these patterns?
Dr. Nemeth ; I suggested four hypotheses. Is it useful for people
to formulate sort of a first guess as to what —
Mr, Ingham ; The only comment I had about your hypotheses was, Abe,
at least two of them had already been answered and I thought the
third one had been agreed upon* The first one, for example, read-
ability, had been answered no. Everybody agreed at the last con-
ference and people agree here that it can be read,
Dr, Nemeth ; This is not a hypothesis. Hieroglyphics can be read
Mr. Ingham ; So you make up a code and then you determine this,
Dr, Nemeth ; No. In other words, when I say "readability", I have
in mind what Dr. Nolan has in mind, tactual discriminability. By
that I mean readability. Can you identify one nine-dot character from
another nine-dot character?
Mr. Strom ; What constitutes identifying a character?
Dr, Nemeth ; By naming the dot numbers.
Mr. Covici ; No, maybe not,
Mr, Strom ; I have a feeling that that is not the best way —
Mr, Covici ; I'm not sure. It's not clear to me that naming of dot
numbers is too significant. It may be that at the beginning naming
the dot numbers is helpful. But, I really don't know,
Dr, Nemeth; May I clarify that?
Chairman Rodgers ; Excuse me. I would just make a comment here and
then I will turn it back to you and pretty soon we will have to sum-
marize this whole thing because time is getting short. You see, this
is the valuable purpose of a "committee". We are not even sure how
to go about these things. So the most we can do is to identify some
of the areas of research,
Dr, Nemeth ; You see, here's what I want to get. I want to get into
the heart of the matter, I want somebody to make a statement. I
don't care whether it's true or false, but a statement as to what we
think will be the result in certain areas of this thing. You can dis-
cuss it forever, and you're not getting close to this idea. But
before' you pan- do research, you have to staj:e a Jiypotjaesis,.^ Maybe it
|Will be borne out, maybe it won't, that I d^n't kno w. That»e thai .
purpose of research. But you also do research this way. Someone
makes a statement at the beginning of a problem that as a result
of this kind of an experiment something will be the case. Then you
do the experiment. Then from the experiment you say, "Well, I was
right", or "No, I was wrong". But you have to begin with a statement.
Dr. Nolan ; I would like to comment. The problem with expressing
hypotheses at this point is that your hypotheses must be phrased in
terms of the actualities of the research situation. This involves
resources, people, apparatus, and so on. How these things will
come into being is unknown to us at present.
I would like to suggest that if we were to only achieve an out-
line of the questions to be studied and these could be broken down
in a variety of ways and hypotheses developed relative to them, re-
lative to testing, this might be a realistic goal. In addition, we
do not know who is going to do this research, and for us to impose
upon this unknown party or parties a format may be a little bit
The other thing noticed is that it is possible to increase,
with a "Nine-Dot Braille" system, the contracting power of a
literary braille and it is also potentially possible to eliminate
some of the perceptual difficulties that result from the fact that
within six-dot braille you have, on occasion, signs that are difficult
to distinguish from one another which require more context cueing
to discriminate between them, I'm thinking in particular of the
lower signs; I'm thinking of the situations where the same sign
stands for three or four different things depending upon positions.
Sometimes where the meanings are particularly unrelated to each
other, like when there is a contraction for "was" or "by". I'm
also concerned with the fact that the so-called signs, like the
italics sign, the composition signs combined with the punctuation
symbols often provide a very large number of cells just for the
process of indicating what kind of character is going on, and
"Nine-Dot Braille" could put, for example, a double capital sign
into one cell and double italics signs in one cell and reduce space
along those lines.
The original "Nine-Dot Braille" had a proposal to eliminate
such extraneous things as the numbers, or such potentially possible
extraneous things, as the number sign, which added one cell every
time you wanted to use a number and also required re-duplication
of the letters "a" to "j" as the numbers. With "Nine-Dot Braille"
the contractions of six-dot braille can be put into a smaller space.
Presumably there is more compactness of presentation on the braille
page. Contractions which previously had half of an empty cell in
them, such as dots 4, 5» 6 followed by a letter, could now be re-
presented in one cell without this additional space, so we have in-
creased the number of contractions; we have more compact represent-
ation; we have reduced the number of ambiguities of other graphic
representation, and we have the facility for making braille capable
of expanding to the needs of the new symbols that have to be created
in special fields - mathematics, music, and dictionary writing. We
have the composition signs made more compact.
Chairman Rodgers ; I think you made mention of all those already
and we have enough material, Mr. Covici, do you wish to add anything?
Mr. Covici ; Yes, I wish to add the following. We have to get around
the six-dot barrier, and there are a number of ways to do this. There
is a need for some improvement over the old system and a nine-dot
system is the thing which should be researched very thoroughly. Then,
if this proves sufficiently worthwhile it should be suggested to
the braille authority, and so forth, as a new standard.
Chairman Rodgers : All right.
Dr. Foulke ; First of all, Mr. Strom, I gather that in the case of
a math symbology, we're not primarily concerned with space saving,
Mr. Strom : No. In any system we are concerned with ease of
Mr. Ingham : Compactness.
Mr. Strom: We're concerned with compactness insofar as it helps
the blind person put more books on his shelf. We're concerned with
compactness in that the finger has less space to traverse in reading
material and, therefore, makes it possible to be done in less time.
John has mentioned here that multiple-cell mathematical represent-
ations are extremely hard to read.
Dr. Foulke : This is what I'm trying to get at. The problem that
arose, then, is that with the braille code it was either impossible,
or else unacceptably clumsy, to express the mathematical language
that you had to express in the braille code.
Mr. Covici : That's right.
Mr. Strom : That's right.
Dr. Foulke ; Then this is a case that you are attempting to make,
Mr. Strom : In the case of mathematics, yes.
Dr. Foulke : That in order to accomplish the job, in order to com-
municate what you needed to communicate, there had to be a larger
supply of symbols available.
Mr. Strom : That's right.
Mr. Covici : Yes.
Dr. Foulke: I wonder if Dr. Nemeth would comment upon that?
Dr. Nemeth: Yes, I would love to. I would like nothing better.
First I wo uld like to make a few a priori judgements as to the space
saving. I think that if the nine-dot system were used, even using
the contractions that Mr. Strom has now, the amount of space occupied
would be from 15 to 20 per cent greater than the amount now required
for literary braille.
Mr. Strom ; Okay. Next a priori judgment!
Dr. Nemeth ; Let me say why. There are I89 contractions in braille.
Short form words consitute, I believe, 76.
Chairman Rodgers ; Yes.
Dr. Nemeth ; Carl, do you know offhand how many initial-letter and
final-letter contractions there are?
Chairman Rodgers ; No, not offhand.*
Mr. Ingham ; I8O something.
Dr. Nemeth ; No, no. Initial-letter contractions and final-letter
contractions only, something in the order of 86 out of 189, "ity",
"ation", and so on. The alphabetic contractions would result in
no space saving. The one-cell part-word signs would result in no
space saving. The only characters which would save space would be
the initial-letter contractions and the final-letter contractions.
Mr. Strom ; Plus any new
Dr. Nemeth ; That's right. Plus any new contractions which you care
to add. I would judge that if you were to use no new contractions
but just Grade II braille, and use the nine-dot cell to represent it,
you would get something in the order of magnitude of about 25 per cent
expansion. But because you added additional contractions in your
code, the expansion will not be quite as great. It will be, however,
about 15 to 20 per cent expansion in number of linear feet that your
hand has to travel to get the information. That's what I mean.
Mr. Strom ; Wait a minute. There was a 25 per cent expansion in the
size of the "Nine-Dot Braille" cell that resulted from the mere extra
Dr. Nemeth ; That's right.
Mr. Strom ; However, you're not counting even the fact that these
39 initial and final-letter contractions save one cell in addition to
*In six-dot Grade II braille, initial and final-letter contractions are
two-celled contractions. There are 33 intial-letter contractions and
1^ final-letter contractions, a total of 4? two-celled contractions.
There are 6? one-celled contractions and 76 short-form words, for a
total of 189 contractions and short-form words (abbreviated words).
Dr. Nemeth ; That's right. Let me say what I wanted to say before
but which Carl didn't let me say, I was saying that on a nine-dot
line you would have 32 cells approximately, maybe 33, which means
you would have to have an average of seven or eight initial-letter
or final-letter contractions per line just to break even. For
example, take the first two lines of, say, the Preamble of the Con-
stitution: "We, the people of the United States, in order to form
a more perfect Union, establish justice," — there isn't a single
initial or fina*l-letter contraction in that statement.
Mr. Strom ; You know this?
Dr. Nemeth ; I was looking through it as I was reciting it. "When
in the course of human events, it becomes necessary for one people" — *
there's the "one" to dissolve the political bonds which have con-
nected them with another " — there's still just the one initial-
letter contraction, "and to assume among the powers of the earth,
the separate and equal station". There's the "ation". There's the
second one, and so on.* You see, you have to have eight per line
to bring up —
Mr. Ingham ; May I interrupt?
Dr. Nemeth ; Let me finish my comment. I will talk long but then I
will keep quiet longer! In other words, as I say, unless you get
the eight initial or final-letter contractions per line, you can't
hope even to break even.
Mr. Strom ; Or new contractions.
Dr. Nemeth ; All right, or new contractions.
Let me tell you my experience with Grade Three, Grade Three
braille, just a brief summary for those of you who do not know, is
divided into seven lines of braille. The seventh line consists of
the modifiers, but the other six lines of braille are the non-modifiers
In Grade Three braille you do this; You put a ^, a 4-5, a 4-5-6, a
5 and a 4-6; in other words, five possible modifiers in front of each
of the six lines of braille. That constitutes 56 characters in the
six lines of braille, and when you multiply this by four, this gets
you 224 initial letter contractions in Grade Three.
In addition to this, there are prefixes and suffixes,
*The word "among" in braille contains the final-letter contractions
There are contractions for "pre" and "pro", and "ate" and "ite",
and "bly", and things of this kind, which are part-word contract-
ions, which are also of this variety, modifier plus principle
I regard myself as a fairly competent braille reader and I
read material in Grade Three, Grade Three is very highly contracted, ,
It has about five or six hundred contractions overall, and you have
to read that system mostly by deduction. What you are essentially
doing is a very rapid mental table look-up, because many of the con-
tractions do not follow normal mnemonic devices. Only 26 of the
characters are letters. The point is, contractions like the "and"
contraction, a certain position turns it into "anyone", "anybody",
"anything". And the "of" contractions of "off", "office",
"official" and so on. In other words, what you are doing when you
are reading it is engaging in a continuous mental table look-up op-
I don't know what the optimum number of symbols is for ease of
learning, and this is a matter for research, of course. But there
comes a time when if you condense the information too much — to
the point where you have to do this consecutive table look-up —
you're going to lose efficiency of recognition. You are going to
gain a lot of space. You would think that since Grade Three is so
highly contracted it would affect the space saving of a proportionate
amount over Grade Two braille, but there is a very rapid low of dim-
inishing returns involved. For example, in Grade Two braille, I
think empirical evidence leads to the fact that the Grade Two braille
effects a saving of about 20 per cent in space over what you would
have if you used simple Grade One braille,
Mr. Ingham ; 25,
Dr, Nemeth ; 26 per cent
Mr. Ingham ; Yes.
Dr. Nemeth ; In other words, you didn't save that much. Now, if you
go to Grade Three by doubling the number of contractions, you add
only a few more percentage points in the saving ratio. In other words,
to effect an additional 20 per cent of saving you would have to add
maybe 2,000 more contraction*.
Mr. Ingham ; There is a good counter argument, but I'll save it
until you're finished,
Dr. Nemeth : Let me finish this. Anyway, this is the thing. There's
a lot of diminishing returns involved here, so that what you gained
in space in Grade Three, you lose in speed of reading.
Now, there's also this: There's no question that the modern six-
dot braille symbol is highly inefficient, and there is a need for
another system. But I think that a revolution in the braille system
that would effect a saving, over Grade Two of only 10 per cent is
not worth the revolution. If you could do it by an order of magnitude
of something like making the distance one half of what it is, that's
worthwhile saving. But just to get a 10 per cent discount at the
cost of such a revolution, I dnn't know.
There's another thing. Let me come to mathematics now. In
mathematics the problem of putting tables into braille would be almost
hopeless in a nine-dot system, for the following reason. It's almost
hopeless already when you have ^0 cells available, but when you have
only 32 columns wide to accommodate some tables, you're in all kinds
of difficulty. You see, you can't put two digits in a slnp^le cell,
can you. Bob?
Mr. Strom ; No.
Dr. Nemeth ; Then you are constricting yourself to a 32-column or
33-column page, whereas 40 or 42 cells is hardly adequate now. oe-
cond of all, I suspect — and I'm making another guess — in mathem-
atics at the level of algebra, in other words —
Mr. Strom : A saving of space over what?
Dr. Nemeth ; Line for line. In other words, every time you write a
formula in six-dot braille and write on in "Nine-Dot Braille", if you
can get one on a line, you can get the other one on the line most
of the time. In other words, in mathematics there won't be an ampli-
ficction. If you are dealing with a special field of mathematics
like vector analysis, where letters are modified by being written in
bold-faced type, then for that kind of mathematics there will be a
saving in the nine-dot system over the six-dot system by a small
percentage. There I would judge it might save 5 or 10 per cent, but
nothing like 20 per cent or 25 per cent at the numbers that I heard
mentioned. I would imagine this is the case.
Mr. Covici : We measured with the -mathematics —
Dr. Nemeth: What do you mean, you measured? How can you measure it?
Mr. Covici t With the mathematics that you have there for the 1962
Conference plus the space taken up by the Nemeth Code, I mean as
far as linear feet of braille was concerned —
Dr. Nemeth ; Wait a while. You see, you are being unfair. You
must not measure a code book. You've got to take a mathematics —
Mr. Ingham; He did it. He took text material and he found it to
be a 40 per cent saving on your code, that is "Principia",
Dr. Nemeth ; I don't know what kind of symbols are in "Principia".
Mr. Ingheim ; It doesn't matter. He did it.
Dr. Nemeth; All right, it doesn't matter. I would say that in
certain branches of mathematics there would be a saving. In no
case would the nine-dot system take more space than the six-dot
system in mathematics as it will in English braille except for table
writing. I don't know what you can do about that. But, as I say,
the nine-dot system will be a saving over the six-dot system in math-
I don't know how you solve this— the problem in mathematics is
not just the proliferation of numbers of symbols. In mathematics,
you have essentially a two-dimensional notation. Fractions are
written above a line and below a line. Superscripts and subscripts
are written obliquely above and obliquely below. Modifiers are
written directly above and directly below the symbols with which they
are concerned. In braille you have, for all practical purposes, only
a one-dimensional method of representation, and the problem stems
more from the need to do one-dimensional representation than it does
from the lack of number of characters. Ken will recognize what I
Mr. Ingham ; I have lots of comments on your arguments when you've
finished. I'll bring them up then,
Dr, Nemeth ; I know you have. Now, for example, my fraction in-
dicators are necessary only because I'm going from two dimensions to
one dimension. In superscripts and subscripts, indicators are ne-
cessary only because I have to go from two dimensions to one dimen-
sion. My alphabet modifiers and my type modifiers, on the other hand,
are necessary because of symbol proliferation.
Mr. Ingham : The problem is, Abe, you bring up so many points so
that when I get back to reply, I can't remember them all!
Dr. Nemeth : Anyway, the point that I want to make is that there
will be a variation in saving mathematics by nine dots over six dots,
depending on the system used, and the saving will be only in the
m-.theraatical -portion of the material, which far outweighs the formulae
Dr. Foulke : Dr. Nemeth, let me ask you a very specific question.
Dr. Nemeth : I h'lve , for all intents and purposes, finished.
Dr. Foulke : As far as space saving is concerned, it seems to me that
again this doesn't have to be cur primary requirement. If a nine-dot
mathematics code will do a job for us that the six-dot code will not
do, we can afford the paper.
Dr. Nemeth : Y e s .
Dr . Foulk e : But what I would like you tn answer specifically then
is, will it do a job that the Nemeth Code will not do or will it do
it significantly better?
Dr. Nem eth : I don't think it will do a job that the Nemeth Code
will not do, because it will also be based on a system of modifiers
just like the Npmeth Code now is,
Mr. Ingham : The real catch to his nuestion is the second part: i-/ill
it do it? Will it do it better?
D r. Nemeth : That I don't know, whether it will do it better or not.
Dr. Foulk e: This, it seems to rae , is what we need to know.
Dr . Nemet h: Better from what point of view?
Kr . Covici : The reader's. I'm not concerned with the transcribers,
Mr. In g ham : Can we hold the answer to the second -oart of the nue.-tion
until I have my points made?
Dr. N ^m^-th: 'fa^t a w^ile. Dr. Foulke h^s asked me a qu'='stion.
Mr, Ingham : I know, but I'm just saying that it would be better if
we c^ulr] hold it.
Dr. Nemeth ; Go ahead, I would like to answer it later.
Mr. Ingham; Yes. I want you to remember it because this is one
of my points also. The first comment is, all of your arguments,
Abe, are based upon the fact that the code in its best form is
based upon the six-dot system.
Dr. Nemeth ; Which code?
Mr. Ingham; The nine-dot codes. This need not be the case. Just
take, for example, your question of algebra and literary braille,
things like that, where you said the savings would be equal or less
than the six-dot braille savings over, say. Grade One, okay?
Dr. Nemeth ; I don't know. I don't follow you; I'm sorry.
Mr. Ingham ; You said that in the literary material you are going
to lose with the nine-dot system.
Dr. Nemeth ; Yes.
Mr . Ingham ; In elementary material in mathematics you're going to
maybe break even.
Dr. Nemeth ; Yes, you're going to break even.
Mr. Ingham; Now, I disagree with you on those points for the simple
reason that there are situations which could be devised, with a
proper choice of symbols, where we would save fantastically. Take,
for example, your chart problem. Suppose your cells, as they often
are, included numbers with signs in front of them, plus and minus,
letters which were capitalized.
Dr. Nemeth ; No.
Mr. Ingham ; Just a minute. Let me finish.
Dr. Nemeth ; You're not realistic at all.
Mr. Ingham; I am so. You can have tables with lots of numbers and
signs in front of the tables. Secondly, you can have tables, partic-
ularly in programing where your letters are all capitals. All IBM
manuals are typed with capital letters. When the transcriber gets
it, she's going to type capital. Now, Mr. Strom can devise a system
where his nine-dot cells include those symbols which in your system
take up a total of four columns.
Dr, Nemeth ; I agree with you for your capital signs, but I don't
agree with you for your plus and minus. There's no way of incorp-
orating a sign for a number with the number in the same cell,
Mr. Ingham ; I don't agree with that. I don't see why Mr. Strom
can't use a group of four dots, say, in the nine-dot cell and use
an extra dot to indicate a minus sign in the absence of a dot plus
a sign, for example.
Chairman Rodgers ; I'm a little confused here and I want a little
clarification for myself. Mr. Strom, have you provided in your
revised code of mathematics for the incorporation of signs of op-
eration with the number itself?
Mr. Strom ; No, not in the current system.
Chairman Rodgers : Do you think it would be feasible to combine a
sign of operation with the number to which it refers?*
Mr. Strom ; It could be, if I were convinced that it would be
advantageous to do so. I don't know whether it is.
Chairman Rodgers ; You don't know?
Mr . Ingham ; This is the question, but it could be done.
Dr. Nemeth ; Ken, do you realize that there are just now the two
possibilities? You have to provide for the possibility of an un-
signed number, a number with a plus sign, a number with a minus sign,
Mr. Ingham ; Go ahead. You still have nine dots to do that.
Dr. Nemeth ; I know, but I have these numbers without —
Mr. Ingham ; Let me just make one point.
Dr. Nemeth ; Let me answer to this one point. Unsigned numbers,
numbers with a plus sign, numbers with a minus sign, numbers which
are preceded by plus or minus, and numbers minus or plus. All of
these possibilities exist.
*In Literary Braille there are special symbols that represent accented
letters in foreign languages, thus combining letter and accent into
one braille cell. See "English Braille American Edition 1959, Revised
1962" - inkprint pages 22-2^
Mr. Covici ; Would the eight cells in the table make a difference?
There are often a lot of tables where they put lines between the
numbers, and they put crazy lines down the page, and they do all
kinds of tricks.
Dr. Nemeth : Let me ask you this: How would you do a simple table
of logarithms: — no signs, just digit after digit? Ten columns,
five digits per entry with a column of "n" on the left, you know, and
across the top? What's involved?
Mr . Ingham : So you use two pages. The thing is this: I agree
with Dr. Foulke that in essence you may lose in certain situations
such as you're pointing out in a table of just pure numbers, un-
signed. Okay, fine. The question is, how would it do with the
rest of the code? I think the point that Mr. Strom made with me over
the phone is that he looked up in the American Physical Society's
booklet, or whatever it was, a list of some 500-odd standard symbols
which they recognized that all print books use, for example, in
articles in their journals. With his system he could account for
every one of them with a single cell,
Mr. Covici : It may not always be possible.
Mr. Ingham : This may not be always true, but the tendency is such
that you can do this.
Another thing as far as your code goes. There are situations in
your exponent problem which is really one of the most important points
in transcribing math, or other sciences, where these level indicators
can be either included in the same cell or can be given a separate
cell of their own, while you use two or three cells in between modi-
fiers. I think all of this discussion is a little bit irrelevant due
to the fact that you will have to admit that a code has to be de-
vised, which can bring these things out much better than what we have
before us here. We really can't say categorically how much it will
save and how much it won't since we don't really know (l) how many
of the signs can be read without ambiguity, and (2) whether Mr. Strom
can devise a code with, say, the logic your code has which, will be
superior to it. I don't think we can give an answer. This is use-
less stuff; that you will end up, say, on an equal basis with
alegbra and you will lose with literary braille. I think that there
are possibilities where you can save quite a bit.
Mr. Covici : And we did save ^0 per cent in the one case.
Dr. Nemeth ; John, could I ask you a question. What kind of symbols
Mr. Covici ; You have the thing there to look at.
Mr. Ingham ; There was everything in it; all sorts of weird
assumptions, and so forth.
Dr. Nemeth ; I don't see it.
Chairman Rodgers ; Go toward the part of mathematics.
Mr. Strom : I wish I had the key for the braille that was used
two years ago, which is completely different.
Chairman Rodgers ; Yes, it is, and this is why I didn't feel it was
Mr. Strom : This is the six-dot braille?
Dr. Nemeth : The six-dot braille. The small sheets are the six-dot
braille. Where am I to look?
Mr. Strom : I better point it out for you.
Mr. Ingham ; I just have one more point to make in order to finish
my 'statement. The point is that ^0 per cent was saved over your code
with this kind of material.
Mr. Covici ; In linear feet of braille. Thirty-six or thirty-
seven per cent; I don't remember.
Mr. Strom ; I can't recall the figure,
Mr. Covici ; It was around there.
Mr. Strom : I don't want to commit myself.
Chairman Rodgers : If I may interject here, we could go back and
forth with the ^0 per cent or no ^0 per cent for the rest of the
afternoon. We don't have the tabulations; we don't have the necessary
material step by step, and a conference like this would not be com-
petent to judge whether a statement of that kind is correct at its
face value and we would have to take it at its face value.
Mr. Ingham : This is part of my statement. There's only one thing
I would like to say. As far as Abe said, he brought up the case of
Grade Three braille, and the saving it made, and so forth and so on.
and I would suggest that much of the problem, as he pointed out
with Grade Three braille, was that it involved the attempt on the
blind reader's part to coalesce two or three cells or say, two
cells, and I think you would win over a system using 500 or more
contractions, as with Grade Three, if they were grouped in a one-
cell configuration and you looked at it as a complete cell^ not as
you insist on doing, Abe, as a cell with a modifier.
Dr. Nemeth r Then you're doing this table look-up that I'm talking
Mr. Ingham ; No, you're not doing any table look-up at all. You're
designing a code which, as you learn it, will be just like the
simple Grade One braille is now.
Dr. Nemeth ; Only 512 instead of 26 letters.
Mr. Ingham ; So what?
Mr. Covici ; When you read it —
Mr. Ingham ; There's nothing mnemonic in Grade Two braille as far
as I can see.
Dr. Nemeth ; There are a lot of things which are not mnemonic. I
don't know where the breakdown begins.
Mr. Ingham ; The point is there's no reason why dots 1-2-4 for
example, should be "f" in mnemonics. There's no good reason for it.
Once you have assigned a characteristic group of symbols to a word
or a part word, there's your mnemonics. It's when you start going
into more bits of information like two cells or more that you con-
tinue to have to look up.
Dr. Nemeth ; What is mnemonic is that "b" stands for "but", and
not for "cow", that "f" stands for "from" and not some other letter.
What you say is correct. These contractions in the code that Bob
has are fairly mnemonic, whereas Grade Three wasn't, but this can
be the case only if you stick to the modifier plus principal chara-
Chairman Rodgers ; I think we will have to terminate this portion of
the discussion. The main point which should have been established
by now, which originated in Dr. Foulke's question was, is there a
need for a "Nine-Dot Braille" system? Correct, Dr. Foulke?
Dr. Foulke ; Yes.
Chairman Rodgers ; I think enough has been said to enable the
participants as a whole to judge for themselves and form a consensus
as to whether there is or there isn't such a need.
Dr. Nemeth spoke of the law of diminishing returns with respect
to space-saving efforts. What I was concerned about, that Dr.
Nemeth didn't mention, and again this is just for the record on be-
half of the average reader, is that the more contractions you put
into a system the more you increase the need for braille rules.
Braille rules up to a point can be tolerated by the average person
so long as they don't become too numerous. Let's suppose that we
have a contraction for "ide" which does not exist in Grade Two braille,
Does it or doesn't it?
Dr. Neneth ; No, not in Grade Two,
Chairman Rodgers : Okay, it does not. In Grade Three it does.
Dr. Nemeth : In Grade Three there is a "de" and you can leave out
Chairman Rodgers : Suppose there's a clearly defined contraction,
the "ide" contraction for "chloride", "confide". That's fine, but
when you come to a word like "confidence"
Dr. Nemeth ; "Mideast"!
Chairman Rodgers: "Mideast", I wonder what's going to happen to
what Mr. Strom has described as the morphology of English words?
Dr . Nolan ; It seems to me again we are way ahead of ourselves.
We're talking about codes and the degree to which they will save space,
I think the purpose of us being here is not to decide these issues
and not to decide whether or not we will adopt "Nine-Dot Braille",
but to decide how it might be studied. And all these things are a
little premature, I believe. They are taking us far afield from
our goal. They are pertinent problems but they are problems that
are going to be an integral part of the study. The codes that we
originate are certainly going to result from consideration of these
problems that have arisen here. But this is for the future. We
don't have the codes jelled now. In fact, there can be several
codes. So wouldn't it be more to our purpose to get back to it?
QUESTIONS ON BRAILLE CODE DESIGN; PHILOSOPHY OF PRESENTATION OF
INFORMATION TO THE BRAILLE READER .
Dr. Nemeth ; First of all, let me put in a plug for modifiers on
the left as being consistent with everything we know about information
theory. Those of you who know my math code I now that one of its
salient features is to give the reader information in advance of
what's going on instead of afterward. So we have the fraction indi-
cators to tell you that a fraction is about to happen, and we have a
modifying indicator to tell you what the precise nature of the mod-
ification might be until later. So we always have to have the
In fact, even in literary braille the people who devised the
Spanish language have an advantage over us. They put the question
mark in front of the sentence to let you know that it's going to be
a question; this is a superior way of conveying information. If you
put modifiers on the end, then you have no readiness for the mod-
ification at all and you are likely to read it in its bald form
before you realize that it should have been modified. That's the
Chairman Rodgers ; Before we go into the second thing, may I
interrupt you there for a second. You do have modifying dots at
the right to close up a situation.
Dr. Nemeth ; To show the termination. They are not modifiers. This
is to put a limit on it. In other words, it tells you when a mod-
ification is ended.
Chairman Rodgers ; Right.
Dr. Nemeth : But this is just so that you could now sort of clear
your readiness for the next thing.
Chairman Rodgers ; Good enough.
Dr. Nemeth ; But the important thing is to know when modification
begins. This is salient. This is paramount. It is important also
to know when it ends, but more important to know when it begins^,.,.
In addition. Bob was making a comment about 8-9 followed by a "p",
I suspect that the psychological confusion — ■-■
Chairman Rodgers ; 7-8?
Dr. Nemeth ; Yes, 7-8. I suspect that the confusion by which it
was read as "er" followed by a trailing dot 4 was more the result
of lack of familiarity with the system than anything else, because
if one is used to a pattern of modifiers plus ^-5i modifier plus
^-5, modifier plus ^-5, one will have a mental set to read the ^-5
by itself and not to initially try to associate the other three
dots with that.
Mr. Ingham ; You see, this is one of the reasons I brought up the
question of the last meeting, and the impression is again in those
notes, I don't think that was a question that came up in the last
meeting at all. In other words, we found that it was readable;
there were no questions as to whether signs were confused with some-
thing else. I think this is an intellectual question that has been
raised rather than a natural practical situation which arose at the
Mr. Covici ; When the dots are on the right in some cases, it may
not be a question of a mental set to read modifier plus body. The
person may still read that more like a "p". Also I think that in
questions of telling you what's going to happen, whether it's on
the left or on the right, whether the modifier should come first to
tell you what's happening, I'm not sure whether the time lag is really
of too much significance; this would have to be researched. I could
see that in a fraction, you would want to know when it actually
begins. That's a whole level space, but whether you have it on the
left or on the right in the cell, it is not a great distance in
space. There's not a very great difference in time. I've seen
material from the last conference. You read a sub-2, say. You read
"a" and you read the sub-2 or most times sub-one.
Dr. Nemeth ; That's a right modifier.
Mr. Covici ; Yes, that's right, but the information doesn't come
Dr. Nemeth : No.
Mr. Ingham ; I think Abe's point was well taken. He was pointing out
that, for example, in the case of literary braille, dot 5 "m" tells
you that the "m" is modified rather than "m" followed by a dot 5.
Yo- don't really read that as a full thing or at least you haven't
where it included two cells for the combination. It may be different
with the one-cell case; I don't know.
Mr. Covici ; Another point about Dr. ^emeth's system; in general
systems of mathematics and of literary braille, there can be at
least two different orientations, (a) the writer, as the person
doing it; and (b) the reader. Dr. Nemeth's code is oriented de-
finitely towards ease of writing by the transcriber. It's very
easy to write and somewhat difficult to read.
One symbol may be strung out over a large area. This impairs
legibility of braille in mathematics notation. The "Nine-Dot
Braille" notation would be more compact. Dr. Nemeth's system is
perfectly computerized. In other words, it would take only a
machine-type operation. But on the other hand, it makes it some-
what harder to read, and I think as far as time is concerned, which
is what we really want to save in the reading, the point of saving
space is to decrease the time; I think it has to be definitely
oriented toward the reader and I think it must be stressed. I
don't think it can be overstated.
Mr. Strom ; I think it is very important to mention that if you're
going to work on designing a system that is in the interests of
the reader rather than the transcriber (which has to be always kept
in mind first) we should hesitate when we say we're developing a
new code for mathematics to mention whether it's absolutely ne-
cessary to have 10 to 15 indicators. Basically, as I understand
it, the present mathematics code gives a good deal of information
on how the imprint shape of the characters may have looked, which
may or may not be important when you consider it in the light of
merely being able to read the material more easily. It may be that
cueing other than that based upon the shape of the imprint chara-
cter is better for developing mathematical symbols.
Dr. Nemeth : May I make a comment about that. There is so much
science around now in mathematics that no transcriber can be expect-
ed to have a comprehension of everything he's asked to transcribe.
Consequently, unless the transcriber can transcribe what he sees,
the likelihood is you get no transcription whatever. In other words,
the ideal thing would be if someone could convey to you the meaning
of the thing instead of the contour of the thing. But it's just
not possible. There are no transcribers that you can get who will
understand algebra, atomic physics, statistics, and what have you.
Therefore, the only safe way to do is to convey notations instead
of significance, and let the blind person read the significance
out of the notation, I'm firmly convinced of this.
Mr . Inghsun ; We're not arguing the Nemeth Code,
Dr. Nemeth ; No. We're arguing the philosophy of presentation. For
example, let me go to the English braille and I'll tell you some of
the things that arise in my mind where this factor has been neglected.
One of the rules of the English braille system says that you should
put abbreviations in front of the number sign instead of after the
number sign. If you want to write "50 feet" you write "ft #50".
This conveys the meaning perfectly well, but you know that a blind
person grows up with a misconception of what that imprint it. He's
likely to do the very same thing on a typewriter and get a grade
for using bad grammar from his teacher who doesn't know what the
rules of braille are. In other words, writing "ft #50" is the kind
of thing that I call a braillism. It conveys the meaning perfectly
well but it does so at the expense of the blind person's full com-
prehension of what the standard situation is in inkprint.
Another thing: A rule of braille says that a transcriber should
translate Roman numerals into Arabic numerals even in references and
so on. So if you want to write II Samuel (Roman numerals), you don't
write the Roman numeral II but you write the Arabic 2. It has come
to the point where I have to have a file card available for my re-
ference as to what the proper way is for writing a reference, because
if I were to use the rule of English braille, I would be accused of
being an illiterate professor. No one writes references like this
in print. And I very often have to use the typewriter. You see, the
blind person's life doesn't begin and end with braille. The inform-
ation has got to come to him from something which has previously been
written and often he has to convey information in written form. So
he has to be aware of both of these, as to what normal print practice
is if he is not to labor under a misconception as to what is right
and what is wrong. The case in point is that I have to keep a re-
ference card as to what the right way is to put a reference in a
footnote because the way you do it in braille is just wrong. You just
don't do it that way. This is the point.
Mr. Covici ; Fine. This is a good point, but is it so bad? I* mean
can a person remember that it's done differently in print?
Dr. Nemeth ; He never gets a chance to find out.
(Dr. Nemeth and Dr. Foulke then proposed some ways and means
of determining the intrinsic degree of ambiguity of certain six
and nine-dot braille characters relative to code design which
were discussed as follows.)
Dr. Nemeth : I haven't done this, but this is a perfectly mathematically
feasible thing to do. Let's go to the six-dot braille cell and let's
imagine that we have before us a sheet of paper of infinite density
so that it is high and wide and we put braille characters on it. Some
of them are truly unambiguous; for example, the "y" or the "sh"
sign or the full cell, but there are certain other cells which have
certain degrees of ambiguity. For example, the letter "a", if it
were present or by itself, you might say has six degrees of ambiguity
because if it were present all by itself you could not tell which of
the six dots it was. The letter "b", you might say has four degrees
of ambiguity because it could be 1-2, it could be 2-3, it could be
4-5, or it could be 5-6. Similarly, you can assign an uncertain
factor to each one of the 6A- braille characters. The space is a .
perfectly certain character, because if there is nothing on the
paper at all, then evidently a space has been printed. The only
question is where!
You can do this and you can sort of weight the amount of uncer-
tainty that there is in standard braille by multiplying each one of
the characters by the uncertainty factor which you have associated with
it, and then you can divide the result by 64 to see how much uncertainty
You can do a similar process with a "Nine-Dot Braille" cell. You
see, when you read braille the uncertainty comes out because of contig-
uity. Things are next to other braille characters or out of context.
For example, a dropped "h" can be either a question mark or an open-
ing quotation mark, depending upon whether it is at the end or the
beginning of a word with which it is associated.
Mr. Covici ; That's a six-dot.
Dr. Nemeth ; That's right. The point it —
Mr. Strom ; Only one side has one meaning.
Mr . Ingham ; Dr. Foulke »s suggestion is that we are talking about
other codes and all of these points should be kept in mind in designing
Dr. Nemeth ; That's right. But what I'm saying is, this is a way
of assessing mathematically the intrinsic uncertainty of ordinary
braille versus the intrinsic uncertainty of "Nine-Dot Braille".
Does everybody understand what I'm talking about?
Mr. Strom ; I understand.
Mr. Covici ; It may be overweighed, overridden by a lot of other
things, you know.
Dr. Nemeth; This may be true, but I'm talking about intrinsic un-
Dr. Foulke ; This can be determined.
Chairman Rodgers ; Hold it. We're getting into cross conversation
Mr. Roberts : Both Mr. Strom and Dr. Foulke wanted to react.
Chairman Rodgers ; I think Dr. Foulke spoke ahead of Mr. Strom.
Dr. Foulke ; I think this could quite easily be done. If I were
going to conduct research of this sort, for instance, I think one of
the things that I would quite likely do would be to give people se-
quences of symbols to read and to cast their results in a confusion
matrix to determine such things as stimulus ambiguity. The procedures
for doing this are quite well known.
Dr. Nemeth ; Yes.
Dr. Foulke ; I think I would go ahead and make the standard inform-
Chairman Rodgers ; I think Mr, Strom spoke next, or tried to.
Mr. Strom ; Yes. One thing that has to be mentioned. I understand
what Dr. Nemeth had to say, this numerical measure of intrinsic po-
tential for confusion is a function of number of dots alone. Is
Dr. Nemeth ;. No, For example, suppose you decide to have nine dots
around the circumference of a circle instead of ranged in a square.
Then the nine dots in the circle would be the same as the number of
nine dots in the square. So it wouldn't be a function of the number
Mr. Strom ; I'm sorry. That's right. But it is a function that's
independent of the assignment of symbolic meanings?
Dr. Nemeth : That's right. I have no intention here of finding
out what meanings you assign to the letters. It's just a measure
of what I call the intrinsic ambiguity of the system - that's all.
Mr. Strom ; Right. Although a more interesting measure, I would
think, would be to take six-dot braille characters and observe with
respect to each six-dot braille shape or configuration how many
different meanings are assigned in that system to that particular
configuration, such as —
Dr. Nemeth ; Take the letter "n". What do you mean now?
Mr. Strom ; Take the letter "n". It has two meanings. It means
the letter "n" and the whole word "not".
Dr. Nemeth ; "not" and that's all.
Mr. Strom ; Okay.
Chairman Rodgers ; Take the dropped "d". I think it has four
Dr. Nemeth ; I know it has several. I think this is true.
Mr. Strom ; Then what you do is you weigh these up and you divide
Dr. Nemeth ; Not by 6^.
Mr. Strom ; I'm sorry. By the total number of meanings that there
Dr. Nemeth ; No, no. You see when you do this you're measuring what
we call contextual uncertainty and then you have to make a statistical
analysis of the trequencjf with which, say, the period occurs, the
frequency with which the "dis" occurs, the frequency with which "dd"
occurs in the English language you know, from large samples, and you
have to weight this. With this kind of a frequency distribution it
would not be fair to say that, for example, the accent mark occurs
as frequently as the letter "e" and use it as weighting the things
equally, you see.
Mr. Strom ; Yes.
Dr. Nemeth ; You have to make a distribution of the frequency of
Mr. Strom ; You are right, and I think this distribution should be
made and the results of this study would be a little more significant
and would be a measure of the so-called intrinsic ambiguity in "Nine-
Dot Braille", which is effectively presupposing that because the
nine-dot cell is bigger, it's possible for the letter "g", for
example, to have one, two, three, four different possible positions.
This will, therefore, tend to increase the ambiguity coefficient.
Dr. Nemeth ; On the other hand, there are so many characters which
are fully wide and fully long, more so than in the six-dot braille,
that the factor will have the effect of decreasing it.
(The discussion on braille-code design then centered on the possibility
of exploiting the full potential of "Nine-Dot Braille", the develop-
ment of experimental codes and the possibility of restricting the
use of "Nine-Dot Braille" to special technical codes only).
Dr. Ashcroft ; I don't know if this is relevant or not, but it seems
to me that some of the kinds of statistical computations around
which questions have been raised today could be attacked right away,
such as the amount of space and the comparative space-saving features
of Mr. Strom's system, and that could proceed right away, and then
I think we could map a strategy, at least a tentative one, right
here for attacking some of the experimental work for a tryout of
some nine dot systems.
Chairman Rodgers ; Of nine dot cell characters; is that it?
Dr. Ashcroft ; Yes.
Mr. Strom ; This would be contingent on developing a code.
Mr. Ingham ; I don't know how you do that unless you're assuming
Dr. Ashcroft ; Let's assume this code for some tentative answers
that we haven't had today.
Mr. Strom: Okay,
Mr. Ingham ; Haven't you gotten that already, any of this? I mean
your figure of M-0 per cent on the math and things like that?
Mr. Strom ; I have not made exact measurements. First of all, this
code and the code of two years ago are different. Furthermore, I
don't really consider either of them representative of the full
potentialities of "Nine-Dot Braille". I think these codes both
represent an extreme construction of the possibilities of "Nine-
Dot Braille", namely that they have been harnessed to certain ideas
which are part of the six-dot concept, namely, the concept that
every contraction must be an initial or final-letter contraction and
must contain a base cell and a modifier in a certain way,*
Also this code has been based upon a principle that anything in
six-dot braille like punctuation marks, and so on, should not be
modified unless we are sure that there is something wrong with it.
So I don't think these two codes are representative, and although
I would be perfectly willing to have various kinds of tests regarding
spacing, and so on, set up with these, I would put in an element of
caution and say that this particular code would not be truly rep-
resentative of what "Nine-Dot Braille" really could be if we gave
a little bit freer rein in designing a code.
Chairman Rodgers ; All right.
Mr. Ljechty ; A question that's been bothering me, Carl, that I
would like an answer to: I don't know whether it's the right place
or not to ask, but I'm going to do so anyhow.
What bothers me is that it is naive, possible to ask: Could
not a technical code be devised on the basis of a nine-dot cell,
that would fit in mechanically as a unit with the standard six-dot
cell, so that you could write a mathematics text using the narrative,
the words in literary braille, making use of the standard literary
braille and simply insert in the line, when the occasion occurs, the
technical sign based on the nine-dot cell? That would permit inter-
pointing of dots on both sides of the paper because your spacing
would be standard and uniform.
Mr. Ingham ; Could I ask for clarification. What do you mean? Do
you mean that, for example, within an equation where you had, for
example, "x" equals "y" square, that "x" would be standard six-dot?
Mr. Liechty ; Yes.
Mr. Ingham ; The equals and others would be nine-dot?
"This is an incorrect description of six-dot Grade Two braille design, sinc«
there are sixty-seven one-celled contractions, consisting of base cell
alone, while there are forty-seven two-celled contractions consisting
of base sign preceded by modifiers,
Mr . Liechty ; Right. That's what I would think, but I don't know.
Mr. Strom ; Let's consider something a little more realistic;
namely, where a line of a formula is all in mathematical braille,
considering that there is an "x" for mathematical braille also.
Mr. Ingham : The reason I brought that up, then, is that if
you're going to go with a nine-dot system for your equations and
formulae, and we're not going to say that the existing English code
is the correct code for "Nine-Dot Braille", then I don't know how
you could represent your letters and things like that, and numbers,
in the two different systems and expect people to be able to read
them or want to read them.
Mr. Liechty ; This is my question. I don't know with codes. So I
don't know how 'to answer, Carl, may I make one more comment and I
think this is the most important comment that I have to make all day;
I'm not so sanguine to think that within our lifetime, or in the
next two or three generations, a basic change from the six-dot to
a nine-dot braille cell is going to be achieved, if it is going to
have to apply to literary braille.
Dr. Scott ; Mr. Rodgers, I wonder if we could ask Mr. Liechty for
clarification of that. Why isn't it likely to happen?
Mr. Liechty ; In my opinion, and this is only an opinion, there
are enough contractions for the average blind reader to learn and
use now. Anymore will defeat any further reading in braille for
the vast majority of people, and this is why I first asked the
question whether or not we can make a distinction and use a new
basic cell for only technical purposes. You will not get John Jones
of Podunk, Iowa, to drop his six-dot cell and go to a nine-dot cell
and learn a new system of braille,
Mr. Ingham ; " I think you are probably quite right in the case of
John Jones, but if you ask me or anyone else in the scientific
fields or music, I might say to you I would be willing to know two
different codes, six-dot and nine-dot.
Mr. Liechty ; I'm sure you're right. My interest in "Nine-Dot
Braille" has been on the basis that this will serve a special people,
an especially selected group of people. It will serve technical,
scientific, mathematical purposes but not for literary braille,
for novels. To read a novel in another system of braille will not
Mr. Covici ; Would it be worth it to you if it increased the speed,
say, by a third? I'm just throwing out a figure of a third,
Mr. Ingham ; I don't think it's a question of that John. I think
it actually will happen, but he's right, it will take two or three
generations for the average guy reading literary braille to move
over, but so what?
Mr. Liechty : In that period of time you would have a duplication
of systems and libraries, and confusion.
Mr. Inghsim ; No duplication of the scientific libraries.
Mr. Liechty : No, but in the literary world.'
Mr. Ingham ; Yes. This may be worth it in the long run. You may
find that some such code turns out to be really worthwhile and to
have a distinct advantage, so that it may take over the literary
Mr. Liechty : I think here then there is a very proper place for
research to find out whether one can determine what is the top limit
of the number of contractions that will be assimilated by the
reader, the blind reader.
Dr. Nolan : I think the value goes beyond that. Maybe nine dot
isn't the answer, but by studying "Nine-Dot Braille" maybe we can
learn more about tactual reading and devise some other system
that may solve our problem.
» * « « i» « ]^
PARTICIPANTS' COMMENTS ON THE DUPLICATION OF "NINE-DOT BRAILLE"
Chairman Rodgers ; Does anyone have any question relevant to the
other pror)osed items of discussion?
Mr. Covici ; I have one question in respect to small and large
Chairman Rodgers ; I think you are referring to the point of de-
termining whether small and large quantity duplication would be
feasible with respect to "Nine-Dot Braille". By small quantity
duplication I meant such things as braille ^writers, braille slates,
and so on. The reason I brought that up, (see Appendix A Working
Paper), is that I think you will recall, Mr. Liechty, at the last
conference, some people were concerned with that aspect,
Mr. Liechty ; Yes,
Chairman Rodgers ; Certainly, duplication of "Nine-Dot Braille"
should be studied in terras of the big IBM computers. But we should
also include the study of duplication by the conventional methods
of stereo,typing equipment, braille writers, and braille slates.
Mr, Liechty ; That's an ordinary mechanical problem,
Mr. Covici ; I think that research on duplication should be done
along with research on code development.
Mr. Liechty ; Still, Carl, you would have the mechanical adaptation
to make of the embossing machine.
Mr. Covici ; Yes, adaptation of duplication techniques and code
development should proceed concurrently.
Chairman Rodgers ; That's a mechanical and engineering problem and
it certainly should be considered.
Mr. Strom ; Mr. Ingham, would you know anything about the possibility
of getting machinery produced so that we could get more of this
material made? I mean I'm thinking of the problem that at the moment
I have a two-line slate. I was wondering whether we could recommend
that we look into the possibilities of getting a machine to do this?
Mr. Ingham ; There is a possibility in this sense that if you could
design a program which would produce the "Nine-Dot Braille" following
your rules from a typist through, say, the IBM machines, it's very
possible that the new MIT high speed brailler could be adapted, I'm
not sure how the keying works, but it probably could be adapted to
produce a heavy-weight paper sample which could then be thermoformed.
The cost there simply would be the cost of the therraoform.
Mr. Strom ; I would like to set up an arrangement whereby this
could be done.
Dr. Nemeth ; Do you want all the same material distributed or do
you want lots of different kinds of material distributed? If you
want fifty pages of the same material, it's not unreasonable to ask
somebody to punch this out on a slate. Don't laugh. Volunteers
d'o this all the time. They transcribe volumes at a time. If you
want to do fifty pages of work, you can do it. I mean somebody
could punch it out in fifteen hours of gross time. Then you could
thermoform it and that could solve the whole problem.
Mr. Ingham ; I know we're all mechanically oriented, but I think
this is a simple solution.
Mr. Strom ; Yes, but there would be errors. For example, I made
22 pages of material for this conference which never were thermo-
formed because of the reading difficulties that we encountered, but
Mr. Rodgers and I looked over it, and there were many erasures that
had to be made.
Mr. Ingham : You have to have an expert transcriber. These are
Dr. Benham- : May I make a point, please. At Haver ford I built a
machine a few years ago that is like the Perkins braille transcriber,
the big thing, except that it's solenoid operated. It's relatively
crude. I made it for the purpose of making legends on diagrams and
things of that sort. It writes on zinc plates. I think for maybe
four or five hundred dollars an extra row of holes could be added
to it to make it nine-dot. Then you could write on a metal plate
Dr. Nemeth ; What about the escapement on it between cell and cell?
You would have to vary it, you know; you would have to make it bigger.
Dr. Benham ; You can make a new rack for that. That's all it is,
just a rack.
Dr. Nemeth : All right.
Chairman Rodgers ; Excuse me. I think we have become too involved
in the point of mechanical reproduction. I would like to ask Dr.
Nolan — I know no one can commit anybody to anything — but, Dr. Nolan,
do you think the American Printing House for the Blind might be in
a position to help with the reproduction of text materials?
Dr. Nolan ; That would depend on several things: One, what we
make them from. Are you talking about vacuum forming? We can
vacuum form all the stuff you want. If you're talking about build-
ing machines to make "Nine-Dot Braille", this is a costly enter-
prise and there might be some questions attached to it.
Mr. Ingham : Can I offer another suggestion. There's another
machine we have in the lab which is what we call a picture brailler
which loads the heavy-weight braille sheets on a rotating drum and
can emboss a 90" x 90" array of dots from recorder or from a com-
puter or from what have you. I'm sure that, for this kind of
density, we could adapt it with a simple program again to do a
master for thermoforming of this type of material. After all,
there are plenty of graduate students and senior thesis students
who could do such a prograim and set up such a thing.
Mr. Strom : Given a description of the machine, I could write the
Mr. Ingham : Fine. I mean this is a simple solution.
Dr. Ashcroft ; It seems that multilith-multigraph approach might
be very easily a solution.
Mr. Covici : There's actually a machine into which you put a
paper tape, and it will make something.
Mr. Strom ; It will operate a tool. Anything that involves progr
ming. I don't know that particular machine.
Mr. Ingham : I was thinking of the flexowriters and machines of that
Mr. Strom ; They have machines to control the operation of the tools
from a program.
Mr. Ingham ; Yes.
Mr. Strom ; So that you could, for example, if you wanted to make
a nine-dot slate with a certain specification of dot spacing, you
could set these specifications up on a paper tape and the paper
tape would control the operation of the drill. That could make a
slate. Again, before I could work with this, I would have to do
things like learning the machine language for that machine paper
Mr. Ingham ; There are minor things,
Mr. Covici; Yes. I just mentioned it
CONSENSUS OF THE PARTICIPANTS ON STRUCTURING AND CPGANIZING AN
EFFECTIVE RESEARCH PROJECT ON SIX AND "NINE-DOT BRAILLE"; REMARKS
CONCERNING FUNDING .
Co-ordinator 's note : At various times during the Conference, the
participants posed questions with respect to the role the Foundation
might be prepared to play in carrying out the participants' recommend-
ations for research on six and "Nine-Dot Braille".
In the interest of conciseness Mr. Roberts' replies have been
consolidated as follows:
Mr. Roberts ; Certainly the Foundation is very much interested in
furthering any efforts that will result in improved systems of
reading for blind people, and I think principally what I would like
to see come out of this is your advice really as to how we can furt^^er
this mutual ob jective . . . . .If there is a role that we have to play and
can play that would De useful, why, as Mr. Barnett said, we certainly
would use our resources to the best of our ability to play that role,
and it is hoped, effectively. If this is to be done alone by the
Foundation, if it is to be done by others outside of the Foundation,
if this is to be done collaboratively, we have no predetermin.ition
on this at this point. We are really looking for your advice and
I should now like to pose a question: What apparatus is there
in this field for launching any kind of organized system of research
that can be identified as a valid proposal here? I should add ( and
this is what I implied before in a more direct response to one question
that was asked) we have no plans for carrying out research in this
area at this point; we have none whatsoever. It may be that we will
and it may be that we won't. I think we want your advice about this.
Let us assume that we identify and agree on areas of research that
are sound and valid. The question I think needs to be thrown out to
the floor as to what the auspices of this should be; what the possible
resources are to undertake this research. We don't know the answer
to this now..., If any proposals come out of the deliberations here
that would seem to indicate a need for this organization--and it does
not necessarily have to be the case — using its resources (dollars,
equipment, staff) to carry out the proposals, we certainly would con-
sider this very sympathetically.
If I just may make this one other comment. I think this has
been implied in my remarks; maybe I ought to state it more directly.
That is, that we have quite frankly no sense of paramount interest or
concern or responsibility in this whole subject that we are talking
about. I think it has been perhaps a long time since this organization,
itself, has conducted research in this area, if it had at all in the
past. I don't even know that,,.,. At this point, I should add, we
have no plah to ask any particular individual or group to undertake
research. It hasn't as yet really been agreed to or identified in
a formal sense.
Chairman Podgers ; Now, I think we are ready to attack the strategy;
in other words, to determine what method of attack we're going to
use to develop a well structured long-range research project. I
think it was Dr. Ashcroft that used the word "strategy", did you not?
And for that reason I would like to start the discussion by asking
Dr. Ashcroft for his ideas on strategy.
Dr. Ashcroft ; It seems to me that we need some kind of a place to
start, whether it's with Mr. Strom's present version of nine-dot or
some modification of it and to get samples of that out for reading,
and to include at that time some of the options that come up to get
these trial readers to feed back options that they see developing for,
say, the assignment of meaning to symbols or introducing more system
to the assignment of meanings. Then to test some of those options in
research and feed the answers back into a revision of the original
version and then we would be ready, it seems to me, for a fairly ex-
tensive trial, experimental trial of the system.
Mr. Strom : Who would operate this? I mean how would such a test
procedure be organized?
Dr. Ashcroft : I think the possibility is limitless there. I think
if you could just get information out about the problem and the idea,
that you will have competent people respond to the challenge, and I
think funds can be obtained to do it, however they would plan to do
Dr. Foulke : If we can tell them what to do, they will do it.
Dr. Nolan : I would not phrase it exactly that way!
Chairman Rodgers ; All right. Dr. Nolan.
Dr. Nolan : I think with the thought I have given to it, what I
might do would be to start out on this legibility thing. I'm so con-
cerned over this. I think Mr. Ingham differs. Maybe some of the
others of you do. Let me tell you why I'm concerned.
In order to get the sensation of touch, the surface of your skin
has to be, for want of a better term, dented. There has to be a dent
made in it. With our regular six-dot cell there's enough space on
either side of the cell so that the skin can droop down, if you will
pardon the use of another non-scientific term, and we can get this
When we go to a nine-dot cell using the same inter-dot spacing
with three dots in a row, the closeness of these is not going to
allow for an indentation to be made by the center dot and this is
going to be lost to many readers.
Dr. Benham : Is that any more true in a horizontal row than in a
vertical row, which we now have?
Dr. Nolan : I don't know. It ought not to be, but you see what
we're running into. We're not actually just talking about rows;
we're going to have rather gross patterns, are we not? We could
have nine dots or we could have eight dots in many figures or even
seven dots. The point of this is until we're sure of the legibility
of these things, to start researching codes may give us problems
because our problem may not be with the code, it may be with the
legibility of the configurations.
Dr. Ashcroft: We have tentative evidence of the legibility already.
Yes I think so.
What do you mean, tentative evidence?
We can read it.
The very reason that you're here is because you're at
a higher level of achievement than many of the people who are going
to be using this thing.
We need to determine legibility for the full range of ability
found among braille readers.
Dr. Foulk e : I don't think we have any kind of evidence of legibility
that we would be willing to act on, do we?
Chairman Rodgers : I think it was Dr. Ashcroft, or was it Mr. Ingham,
who mentioned that we have tentative evidence.
Dr. Ashcroft : I did.
Chairman Rodgers : I would like to hear it Dr. Ashcroft. Will you
put it on the record.
Dr. Ashcroft : One of the earliest comments I heard at this conference
was that even the crude material that you had last time, which had
spacing problems in it and other problems, was read by people here.
Mr. Ingham : I think this is probably derived from my comments. I
mean I'm no protagonist for "Nine-Dot. Braille" ; I'm willing to go
either way. I still think we should do research.
Dr. Ashcroft ; I think there are two approaches to the problem of
legibility. One is a gross one where you just provide people some
stuff and let them see if they can read it, or the analytic one that
Dr. Nolan is suggesting, and I would certainly go to the gross one
before the analytical one.
Dr. Foulke : I don't think I would. If I were going to start re-
searching this problem, I would conduct, I think, an initial study
to try to determine or learn something about the legibility of all
of the characters that were possible in this system, and I think I
would probably want to set up at least some kind of tentative criteria
for symbol legibility, and on the basis of this then attempt to form-
ulate some kind of code.
Dr. Ashcroft : Then you will have to do the legibility studies on
symbols in isolation and you won't know the answers to whether it
can be read or not.
Dr. Nolan : Wait a minute now. Let's just go a bit further. I think
a check on legibility should certainly be made. This doesn't mean
that everything has to stop while this is being done. It may be that
the initial efforts on legibility will show that this is really an un-
necessary problem. It may be that we will see that we really have
legibility problems and if this is the case, then it's rather reck-
less at the moment to get involved in codes until we solve this problem.
Mr, Covici ; What do you mean by legibility, counting how many dots
there are or recognizing the difference?
Dr. Nolan : There might be a very elementary way. You might dis-
criminate different patterns or you might just report the stimuli
present. There are several approaches that you might use.
Chairman Rodgers : Just a moment. I wonder whether the two phases,
the legibility phase and the stimulus phase, can be investigated con-
Dr. Nolan : What I think I would do if I had this problem to solve, I
would work on legibility. I would start exploratory work with the
code Mr. Strom has devised. This would be a trial and error process —
actually trying to accumulate sources of error in the code. While I
was doing this also I would be refining the code on a logical basis
and also working into these special areas where special codes need
to be developed. Then at the same time, or a little later, I would
just for the heck of it develop a whole new literary code and study
this to see how we rairht alter the degree to which we can communicate
in tactual form.
Dr, Ashcrof t: By new literary code you mean six-dot or nine-dot?
Dr, Nolan : Nine-dot.
Dr. Ashcroft : How would you differentiate what we've been talking
about from literary code?
Dr. Nolan : The present setup is to use the old literary code as the
base and modify it, isn't it?
Mr. InRham : That's right.
Chairman Rodgers : Yes.
Dr. Nolan ; We know, on the basis of our experience with the present
literary code, somethinr^ about soft spots in it. For example, fre-
quency of occurrence of letters, etc., are not reflected in the pre-
sent code. I would develop a new literary code on the basis of what
we know about information theory, what we know about the problems we
have found in your present literary code, and so on. Remember, vre're
researching, we're studying, we're not saying that we're going to
adopt any of these things, that any of them will ever c-me into prac-
tice. But where are we now? Here we've had this problem for years.
We have "trialed-and-errored" it through. No one has ever extensively
explored beyond the present system; what has occurred has been more
a fight and a power struggle than an exploration. Here we have an
opnortunity to take off in several directions at the horizon. What's
over it, we don't know. Let's look and see!
Dr. Benham ; It seems to me that there are few enough people reading
braille as it is, and the harder we make it to read the fewer there
will be. I don't know what the percentage drop in reading was when
it went from Grade One to One-and-a-half and then when it went from
One-and-a-half to Two, but I'll bet it drops every time there's an
increase in the complexity of it. So I think that Dr. Nolan's
suggestion of finding out the legibility of the system, nine-dot
system or any other system, is extremely important because, yes, we can
I agree that at the first conference we had, I could read the
"Nine-Dot Braille" that I was presented with and, in fact, I believe
I even wrote a little of it on the slates. Sure, I could do it. But
it was a contest; it was a game. I was vying with my fellow man
trying to be as clever as he was and he was trying to be cleverer than
I was,. and as a result we were stimulated sufficiently to manage to
struggle thought it. But I imagine I'm the poorest braille reader
here today. Because I can read braille only when it's absolutely
necessary that I read it, like for instance, the Working Paper that
we were presented with. I read through those pages, but it was a
struggle. I don't know why this is so; I wish I did. But I really
get frustrated trying to read braille. I don't know why it is. It's
obviously some psychological thing in my background some place, but
that happens to me,
I can struggle through mathematics and physics and engineering
and science of all kinds of that nature because it's a slow, tedious
process reading it anyhow. So it doesn't bother me. But to pick up
braille and read it as a novel, I just wouldn't think of it, I
couldn't. I can't even read the braille technical press (and I'm
interested in what's in there), partly because there is so much re-
dundancy in it. In your working paper, for instance, I nearly
through the thing out the window when I got to the last couple of
pages and you repeated and repeated and repeated in each line. Why
you couldn't have said it at the top of the page and then given the
salient features of what it was you were doing, but every paragraph
is just like every other paragraph except there was one number
difference in it. And I had to read through the whole thing just
to find that different number, and I was afraid to skip it for
fear you might have slipped something in on me.
Chairman Rodgers : You're now talking about the format of the
braille; is that it?
Dr. Benham : No, I'm just talking about the fact that it's hard
for me to read; it's slow, tedious, and I read it only when I have
to and, of course, I have to, and I have to quite a lot. But I don't
do it for fun. And the Grade Two contractions throw me every once
in a while. I'm trying to read along and I come to some cockeyed
looking this. I can't make out whether it's capital "St" or "each"
or some other combination of dots, I'm not being very coherent,
but then I'm not intending to be.
The point that I'm trying to make is that I think that when
we go to something that is harder to read than what we already have,
we'd better be sure that it's going to be acceptable to, I would say,
at least as many people as now find Grade Two acceptable, because
we're going to lose people on the way. Another interesting comment
that I have often though of is: I wonder whether blind people are
poor readers because they never see words spelled. Everything
you read in braille is a contraction of some sort as the "m" for
"more", but you never see "more" spelled out. We're going to get the
system so contracted that its going to get even worse in the literary
area if we're not careful. I think I told you at the beginning that
there was a great deal of redundancy, because I agree with a great
deal of what's been said, I was only trying to emphasize a few
points. For literary braille, I'm not sure that we want to make the
thing more contracted than it is, more difficult to learn, and more
difficult for some people at least to read. We ought to try to make
it easier to read, not harder,
I find also that the "Nine-Dot Braille" that I've seen so far
is too spread out. For instance, when you take the material that
you presented us with today, the code, and you look at the nine-dot
version of it and then look at what it means, like "pro" or whatever
the letters are, those letters look like they were sitting in the
middle of the desert,
Mr. Covici ; You mean the six-dot braille.
Dr. Benham ; Yes. But obviously you're going to have to spell
some things out. You're not going to have everything abbreviated,
I hope. So there will be occasions where you want to follow one
letter with another. What is it going to look like when you do
that? Is it going to be spread out? Now, I've been inconsistent
because in some cases I say you're getting it too compact, and in
some cases I say you're getting it too spread out, I think it
boils down to the fact that I believe the present braille spacing
in the six-dot code is about the optimum spacing for easy reading.
To get it much closer, you have to study it too much except for
the exceptional braille readers, and if it's too spread out it gets
difficult to read and you can get lost in it. This is one of the
things that I dislike about the Nemeth Code. It's too spread out
on the page as far as I'm concerned. Of course, I was brought up
on a different one, so my opinion is really not important, but
for me it's too s-oread out in many cases. I want to find out what
the legibility of a new system is, before we go very far in trying
to decide what kind of a code it ought to be and what kind of ab-
breviations there ought to be.
Dr. Ashcroft : I was much in accord with what Dr, Nolan said about
carrying on legibility studies, but I think some kind of actual
reading study, some kind of a nine-dot code, should go parallel with
Dr, Nolan ; Yes,
Mr. Ingham ; I agree.
Mr. Covici ; Yes,
Dr, Ashcroft ; I think this is in line with what Dr. Benham has
just said, too.
Chairman Rodgers ; Yes, I think so, and perhaps the two studies
could go on concurrently, but I would like to comment in particular
about what Dr. Benham has said with respect to reading difficulties,
I think that in the study of legibility, one thing we have to be
very careful about is not just whether the finger can discriminate,
but how readily? In other words, there is a difference between
mere discrimination and getting by, and genuine reading; and that
is, I think, what we're concerned with,
Mr, Covici; It would take training.
Chairman Rodgers : Yes, it would take training. In other words,
one should be able to discriminate without too much study of the
thing, you see. That's the point I think that should be very care-
fully taken into account in any study of legibility. Maybe I mis-
used the words "degree of discriminability" . I should perhaps have
said "discriminating without too much studying of what you have
Mr, Strom ; Then, Mr. Rodgers, would you say that the experiment
should go along the lines of a point where we spend several months
training a certain number of subjects to become proficient with
"Nine-Dot Braille", and then make measurements with respect to how
quickly they pick them up?
Chairman Rodgers : Yes. I just tried to throw that in about ready
discrimination insofar as helping the study is concerned. But as
to what specifically to do, that's a matter of structuring and that
is what we're here for, to collect your opinions as research special-
ists. That should be up to you to decide and if anyone would care
to comment on Mr. Strom's suggestion —
Mr. Ingham : What was the suggestion again?
Mr. Strom : That among the tasks that we consider would be one in
which subjects would be trained for a few months to learn the code
and develop proficiency in adapting to its peculiarities, and then
train them on being able to read straight text of the material and
see how easily they are able to,
Mr. Ingham ; Which code are you talking about?
Mr. Strom ; Either the current "Nine-Dot Braille" code or one of a
few modifications thereof. I think the point mentioned about two
"Nine-Dot Brailles", namely, one like the one here and another one
that we invent on the basis of six-dot braille but, rather, on the
basis of information theory, optimization of whatever factors we
figure will make it more psychologically readable. I think we should
try, perhaps, working with both of these, but the important thing
is that we have got to get some kind of tests made with some code.
The more codes we get tested, perhaps the better off we are.
Mr . Ingham : What Dr. Nolan was saying requires that you people
go into lots of practice. If you're going to use a new one, you
have to have lots of practice, of course, i
Mr, Strom ; Then we have to first of all get people in our project
who are specializing in teaching braille, I mean among the people
who would carry out a research program will be specialists in
education, people who are familiar with the specific problems involved
in teaching of braille and who would be able to adapt this to
"Nine-Dot Braille" so that if we have a research group, one of the
members of the group would include a specialist in education.
Dr. Scott ; Have you, Mr. Strom, at the present time, developed
any code which exploits to its maximum the potential "Nine-Dot
Mr. Strom ; No. I haven't.
Dr. Scott ; Would this be a relatively simple thing to do or
would it be time consuming?
Mr. Strom : It would be relatively easy to invent a code that
broke away from the six-dot braille and gave more flexibility.
What would not be relatively easy would be to optimize certain
details within this.
Mr. Ingham : You actually could, particularly if you could take
advantage of, say, the lists of the ^00 more frequent words in
English text, and so on and so forth, which the English braille
Mr. Strom : Yes, but also what I have to take into consideration
are the 512 nine-dot configurations and some kind of measure of
the degree of appropriateness of these contractions. So you have
a whole bunch of lists of common words or word parts and then you
have a list of 512 dot configurations. We have to somehow isolate
a property of these orthographic configurations which we can relate
to a rule for assigning proper meanings to them.
Dr. Foulke : This you can probably do on the basis of linguistic
information, for instance, construct a code taking into account
such things as sequential dependencies, orthographic conventions,
things of this sort, but in addition to that, before constructing
a complete rational code, it seems to me that you would have to
know something more about the stimulus alt)habet itself and we're
back to legibility again.
Mr. Strom : That's right,
Mr, Govici : Can't you do both things at once?
Mr, Strom : The problem is I could write a tentative code but I
could not write easily an optimum tentative code. What would have
to happen is first an intuitive code and then give it a legibility
test; train subjects to read it, assess their reactions, and get
verbal questionnaires from them, get statistics on speed and, in
short, get quantitative and qualitative data on this. Then have
the specialized studies regarding problems of individual characters
independent of meaning, as somebody suggested. Then we have these
two studies. These have to be correlated.
Then there has to be a feedback mechanism where the code is
revised and -out through the same tests, so that you are always
using information that you are gaining about psychological prop-
erties, legibility of characters relative to the meanings you
assign to them, and you have to use this information and put it
in the context of a whole system which you then have to put to the
tests. The tests will take a long time. Every time you have a
new code you have to get some new subjects and train them for
several months on this. Presumably these tests would have to be
made — you would have to get a whole group of blind subjects.
Mr. Covici : Maybe we could have the subjects optimize the code
by these tests, maybe change it a little bit and make it better
and then eventually we can get enough legibility studies through
and then maybe eventually we will have one code. Then we will try
it on a whole group of people and get your statistics back.
Chairman Rodgers ; It seems to me that that is a statement that
the research specialists would have to approve; that is, in terms
of established norms of research procedures. Whether it would be
a sound one would be up to the research specialists to determine.
Would any one of you care to comment on Mr. Covici 's remarks?
Mr. Ingham : I think it's haphazard. That's no way to go about
doing it. Could I interject a hopeful note as far as one of the
things that was brought up. One of the hardest things I had to do
was to find how many people do read braille. If anybody has a
concrete fig:ure, I have, and this is somewhere between 30,000
and 40,000 who are reading braille at the present time; this number
is constantly increasing and has never declined actually.
Dr. Nemeth : What constitutes reading braille? What material do
Mr. Ingham ; I don't mean just using it for notes; reading books.
For example, more concrete figures in that line would be 4,000
people subscribe to the braille book listing. Somewhere between^
5,000 and 6,000 read the Reader's Digest, and things like that.
These are just figures, though, in case you're interested. So
the figure is increasing. Tnis will be an extremely useful thing
to have if it will work.
Chairman Rodgers : Let us now get back to the structuring of the
project. Dr. Benham, I think, has given us his views. Dr. Foulke,
did you have anything to add to the structuring of the project?
Dr. Foulke : No, I don't think I have anything to add. I think I
could summarize what I would do if I were going to research.
Chairman Rodgers ; All right, do you want to give us your summary.
Dr. Foulke : I think I would start out with a legibility study in
an effort to learn something about the legibility of the dot com-
binations that are possible with the nine-dot system.
I think I would attempt, on a gross basis, just to learn
something about how people would respond to a system such as the
one that is presently before us; that is, the attempt to teach it
to some people and get some impression about how well it did appear
to work, and then I think, probably at the same time, I would
attempt to construct a brand new code on the most rational basis
possible, using all of the combinations that are available within
a nine-dot system on the basis of an analysis of the language, re-
dundancy, of sequential dependencies, things of this sort and what
was learned about legibility or ambiguity of the elements in the
stimulus alphabet. Also I would give some consideration, I guess,
in the assignment of elements in the stimulus alphabet to elements
in the response alphabet,
I would give some attention to the matter of stimulus-resrionse
compatibility. I think that this would be the kind of research
program, then, that I would undertake if I were going to do it.
Chairman Rodgers : Thank you. Mr, Ingham.
Mr. Ingham : As far as the projects go, I think what has been said
just about summarizes what I would like to say, except that I would
reinforce the comment about a gross survey, about, for example,
having the present code used to do maybe fifty pages or a hundred
and distribute it to a reasonably large number of people with
questionnaires, carefully designed to elicit comments on both chara-
cters and context and understanding. I would do this as soon as
Chairman Rodgers : Mr. Liechty, would you care to summarize your
Mr. Liechty : The notes that I've taken, I think I've already touched
on them. I think I will pass up and save the time. Thank you.
Chairman Rodgers ; Dr. Nemeth, what would you suggest with respect
to research procedure?
Dr. Nemeth ; I would suggest that we should undertake several
directions of investigation: One, produce just ordinary English
braille with no additional contractions but just the contractions
of English braille in the nine-dot cell form, and allow people to
accustom themselves to reading just that. You realize that they
will thereby only need initial and final-letter contractions in
this form of the research. They will not have to learn anything
new essentially but just to acclimate themselves to the change of
the spacing, and to find out what kind of acceptance or what kind of
troubles you get there.
Two, I would make a statistical analysis of the frequency of
occurrence of words in order to determine what kind of codes
would be most suitable for them. I think that a one-to-one assign-
ment of 512 nine-dot sub-sets to the 512 most frequently occurring
letters or groups of letters would just be fatal. I don't think
you would get anywhere,
Mr. Strom ; I agree. It has to be different.
Dr. Nemeth ; So this is the second thing. Then I would suggest
that the thing to do is to bring the code that you eventually devise
gradually to bear, produce material maybe a volume long, and on
the first two pages you do nothing but Grade Two, and then you
add three contractions of the code you devise on the next two pap:es,
and then you add three more contractions, and so on, because learn-
ing the code and reading the thing all at once is going to get you
an unrealistic response,
Mr. Covici ; What if some of the contractions are different? I
mean you want to keep all the Grade Two contractions in everything
Dr. Nemeth ; No, For example, the contraction for "in"--Bob made
it three dots wide to avoid the other troubles. The contraction
for "can", he also made three dots wide for the same purpose.
Mr. Strom ; So that it could be a part-word sign, that you could
use it in the middle of a word without its looking like a letter.
Dr. Nemeth ; I see. I thought it . was for orientation. In other
words, vou would not wonder whether it was "a" followed by dot 1,
or whether it was "a" followed by dot k,
Mr. Strom ; It serves both purposes.
Dr. Nemeth : Anyway, there are relatively few changes. The "it" is
another change which can be used as a part-word sign. In other
words, you could introduce all of Grade Two plus those contractions
of Grade Two, which are changed and then add brand new ones as you
go. I'm afraid that while it will be interesting to make the kind
of experiment that Dr. Nolan suggested; namely, the ease of perceiv-
ing meaningless dot configurations, I don't think it does begin to
tell the story of readability.
Mr. Covici : We do not know.
Dr. Nemeth : That's all I have to say.
Chairman Rodgers : Thank you.
Dr. Nolan : I can only say I disagree!
Dr. Nemeth : Okay.
Dr. Nolan : I have already given my comments on the wav I think
the research should proceed.
Chairman Rodgers : Yes.
Dr. Nolan : I don't think I have anything more to add. -
Chairman Rodgers : You don't have anything else?
Dr . Nolan : Only to say that I disagree with my friend, Dr. Nemeth.
Chairman Rodgers ; I was just going to interject a short comment
which may be a redundancy. As I see it, readability is to a large
extent a highly ];efined form of discriminability , so that when we're
studying degrees of discriminability for the purpose of relating it
to legibility, we have to bear that in mind very carefully.
Dr. Nolan : That's right. I would go along with that, but legibility
is a basic ingredient.
Chairman Rodgers ; That's it,
Mr. Strom; It's one factor among many.
Chairman Rodgers : Dr. Scott?
Dr, Scott : I want to make two brief comments, but I want to
preface them by saying that I know nothing about braille and so my
comments are to be taken with that in view.
One is that I have the feeling we don't really have a great
deal of research on the six-dot braille cell and this is one of the
difficulties we have had here, because as questions come up about
"Nine-Dot Braille" it becomes apparent that we don't even know the
answers about six-dot braille, and I think it would be a mistake
if the research strategy were to study only "Nine-Dot Braille",
because at the end of it we would only know whether or not it
was a feasible system and if it weren't a feasible system, we would
have gone through a lot of trouble to determine this. It seems to
me a much more realistic thing to clue in the basic problems about
which we have no data at the present time which are applicable
not only to the six-dot but to the "Nine-Dot Braille", as well.
In other words, we ought to be studying both systems or at least
our research ought to be phrased so that it's applicable to both
The second thing is that there is a analogous experience
which the field of education has recently gone through and I imagine
that we might learn something from this experience, because there
has been an attempt to introduce a new system of reading involving
various contractions and this is being experimented with by the
public school systems*
I don't think we will learn anything in terms of how to set
up a new system of raised-dot reading, but we might learn something
about how you go about instituting it or about what kind of -problems
you meet in doing so, or, if it were decided to use "Nine-Dot Braille",
I'm sure that the experience of these people trying to institute a
new system of reading would be very valuable, and also how they went
about researching it, which would be a very valuable thing in terms
of setting up strategies here.
Chairman Rodgers : With respect to that research, of course, I
think it could be very helpful to us in obtaining all the reports
on the methods and results thereof. So we're very happy to have
Now Mr. Strom.
*(See NEA Journal , "Value of ITA", September 196^ - Pages 20 to 22).
Mr. Strom ; I go along with the proposals of all the people who
have mentioned most of the things to be considered already. The
very fact that we have disagreements in matters of details with
regard to research strategy, implies that we somehow have to set up
a structure of a small group of research assistants who are going
to plan this, and it is somehow going to finally come to a con-
clusion regarding different ways of going about the problems and
actually putting them into execution.
The group has to be small but diversified enough so that it
contains representatives from the fields of psychology, education,
braille, mathematics, and whatever other specific categories are
necessary, and who are fsuniliar with matters regarding machinery
and with natters regarding testing methods and educational methods.
Further, this group has to be run sort of semi-independently and,
therefore, we somehow have to recommend that a small sub-committee,
say the braille legibility committee, the braille legibility testing
or something like that, be set up as a result of this conference
to ascertain what kind of equipment and research is necessary and
to plan this program and then perhaps solicit funds from the
American Foundation for the Blind for carrying this out.
Some small group would need to give basic direction for the
research efforts, and I think now is the time to decide how we're
going to go about getting such a group.
Mr. Covici : The first thing I would do in a research project is
to review all the literature that's around and find out how valuable
it is. I don't think anybody has mentioned this but I think it
should be done, even though I don't know how much there is.
Secondly, I would set up an autonomous research group. In
other words, I agree substantially with Robert's statement, but I
must emphasize that it should not be too closely linked, it should
not have too many administrators on it. It should be research
people in different fields who could do a little work independently
of the group as well as with the group, and so forth. I would set
up studies; I would develop a code; actually develop the material,
have people read it, and I would also do some legibility studies
on the meaning of the combinations just so that some interest
might come out of it. I would do some information theory, and so
forth, to develop a code and in general to do what he said, but
I really must emphasize that we've got to get something out and
have people read it. What good is it if they don't?
Chairman Rodgers : If no one else has any comment — and the time
is gettiniT a little short — I should like to get to the final phase
of the Conference; namely, the organizational aspect of the structure.
I thoroughly agree that it would be most unfortunate if no action
were to result from this second Conference, because this is the
very reason we called a second Conference on "Nine-Dot Braille".
I, therefore, am open to suggestions as to what steps should be taken
to organize the nucleus. I personally feel that this is a highly
competent group, and think you all could make valuable contributions,
if you were to organize yourselves into a research nucleus from
which the whole project would stem and develop. By "whole project"
I mean the research contents which are now on record.
You may disagree with that proposal and we will mnke the
rounds again in the regular alphabetical order. Dr. Ashcroft,
what do you think about this nucleus idea and about this group of
participants constituting it?
D r. Ashcroft : I think that this group might constitute an advisory
committee. They might be very helpful, but I think ultimately
the work will have to be done by individuals, and I would like to
see this group or some group that is constituted for this purpose
of being an advisory committee publicize the work widely and
attempt to get the individuals interested who need to be. And then
facilitate the writing of proposals by the individuals for funding
by private agencies like the Foundation, although I'm sure that there
are also federal funds that could be solicited through grant t)ro-
posals to carry on the work. But it seems to me that ultimately
the work has to reside in the hands of individuals to get it
Chairman Rodgers : All right. Dr. Benham.
Dr. Benham : Whom did you put down on that nucleus?
Chairman Rodgers : All the participants who are research specialists.
Dr. Benham : Then I have no other comment. i
Chairman Rodgers : Dr. Foulke, do you have anything to suggest or
add to or modify?
Dr. Foulke ; If we are to set up an advisory committee, I'm sure
we'll all be happy to advise anybody we could get to take it. I
don't know if this is the way in which we are to serve as a con-
sequence of this meeting today, I'm really uncertain about what our
next step would be. If we were to serve as the group who is actually
to divide up the problem and define missions and get some research
on the way, there is going to have to be some money to do it,
Mr. Ingham ; I think you will have something if that's what
you're getting at, if you're telling us to go look for somebody »
I don't know about that. Let me ask a question. There seems to be
little doubt in my mind that Boh, for example, could polish his
code along the line that we suggested. The question is suppose
he did have a code? It's also pretty apparent that we could find
a way of getting at least a sample prepared of varying material.
Okay, Now, can anyone in this committee take on the job,
with a certain aunount of financial support, say, from the Foundation,
to set up a reasonable research questionnaire, or whatever have you,
thermoform a very long set of copies, and distribute many hundreds
of them, carrying on first this gross kind of test, and maybe per-
haps somebody else in the committee here will be willing to go on
at the same time carrying on fine legibility research? These two
approaches immediately offhand could be quickly carried out, I
think, and at a minimum of cost, I just wonder whether we could
start on that now. Do we need these as definite avenues to work
Chairman Rodgers : To whom did you address that?
Mr. Ingham ; I'm addressing it to the Foundation or the powers
that be or anyone who would like to take the ball as far as trying
this research out.
Chairman Rodgers ; As far as the Foundation is concerned — and Mr,
Roberts is not here now, so I have to speak on behalf of the
Foundation — again I have to fall back on what Mr, Barnett said
this morning. It's on the record. If that is the consensus, if
that is the recommendation of this group, the Foundation will try
to do all it can to help.
I seem to sense that if the problem of funding is solved
either through the Foundation or through the United States Office of
Mr . Ingham ; Or anybody.
Chairman Rodgers ; Or anybody else, that this group is ready to
take on specific assignments with respect to carrying on the
Mr. Ingham ; Why don't we go around and ask everybody that question?
I think that would help.
Chairman Rodgers ; I certainly propose to do just that.
Mr« Strom : Everybody isn't here now,
Mr. Liechty ; Dr. Nolan had to leave, Carl, and he sent his regrets
to you. You were talking and he didn't want to interrupt and he
said goodbye. He has to catch a plane. Dr. Ashcroft stepped out
into the library and will be back in a few minutes.*
Chairman Rodgers ; All right. Then we can proceed in the meantime
with the others, and the first one would be Dr. Benham.
Dr. Benham : I don't feel qualified to do more than heckle in any
research program that might be set up, because there are a lot of
the aspects of it, psychological and so forth, that are not down
my line. I could offer assistance in the way of perhaps getting
some hardware taken care of, or distribution or names of people
that you might want to send the material to and things of that sort.
But as far as taking any major responsibility in heading up or in
working out the research program, I don't think I'm qualified. So
I will offer my services wherein they are useful, but it wouldn't
be in the top echelon part of it.
Chairman Rodgers : By "hardware" you mean reproduction of the
material; is that it?
*Dr. Nolan stated in his subsequent letter of June 22nd, 19^4, as
"Please accept my apologies for having to leave last Friday's
conference before it was over. I was especially sorry not to be
able to participate in considering how the research might be under-
"I have since talked with Emerson Foulke and he described the
statements of intention to participate made by the members of the
conference. I certainly feel that APH would be willing to contribute
by reproducing materials and possibly by making computer time avail-
able for data analysis. It does not appear to me that individual
and cooperative efforts of the type agreed upon are the best way of
energetically studying this problem, A better way would be to set
up a funded project with its own personnel for just this purpose.
Project staff could, of course, rely on outsiders for assistance of
various kinds. This could be done at AFB or through a nearby
university and would provide continuity that is rarely possible
through cooperative efforts."
Dr. Benham ; Yes, or getting machinery of one kind or another
that you might want. Perhaps even being able to find a little
money here and there, but not from the point of view of setting
up an actual research program.
Chairman Rodgers : The design for the research project itself.
That in itself is essential to carrying on the work.
All right, that's on the record. Dr. Foulke, would you
be agreeable to accept an assignment granted that the funding
problem was solved, and so on?
Dr. Foulke ; Yes. If funding were available, I would be willing
to participate in any of this work, and if I had my choice I
would prefer to work on an effort to develop an ideal code based
upon what information is now available and could be collected
within a reasonable amount of time.
Chairman Rodgers ; Dr. Ashcroft, the question on the floor now
is being addressed to each individual participant, and the question
is: Assuming that the funding problem in carrying out a well
planned long range research project were to be solved, either
through the American Foundation for the Blind or the United States
Government or a combination of private and public dollars, would
you be willing to serve insofar as contributing to the project?
Dr. Ashcroft : Yes, perhaps primarily indirectly through scholars
in our own program at Peabody. A direct contribution on my own
part would be secondary to the commitments I have, but I would be
interested in participating on both ways.
Chairman Rodgers ; I see. Then you would be in a position to
participate actively in some way. In what particular phase of
research would you be interested in p^^rtici pating?
Dr, Ashcroft ; I think primarily at the applied level of actual
trials of nine-dot codes, field testing, and perhaps if there is
a place for it, especially in terms of children and the introduction
in educational programs. One other thing that I would be interested
in is a programed instructional approach to moving "Nine-Dot Braille"
into some kind of use.
Dr. Nemeth ; What approach? I didn't hear,
Mr. Strom ; Programmed instruction.
Dr. Nemeth ; I see.
Mr. Strom ; Perhaps we could get some people at IBM to help us
out on this. I know a few people there who are interested in braille,
and in programs.
Chairman Podgers : Of course, Mr, Covici and Mr. Strom, they would
not want to participate at all! My point is that I think we know
what your roles would be.
Mr. Covici ; Yes.
Ch-=iirman Rodgers ; The development of codes and so on, right?
Mr, Strom ; That's right. Also as far as programming computers
are concerned and that technical area. I mean things like results
of psychological experiments, program, production of "Nine-Dot
Braille" according to certain transcription rules that can be fed
into any machinery that Dr. Benhara or Mr, Ingham can get for me, I
can work on.
Mr. Ingham ; I've already indicated that I am very much interested,
both directly and indirectly, through whomever we can grab at the
lab at MIT. I can't commit them, obviously.
Chairman Rodgers ; Dr. Nemeth, are you a research specialist?
Dr. Nemeth ; Well, I've done research in the past. I would be
willing to do this much. I would be willing to heckle along the
literary lines, but when it comes to the mathematical end of it,
I would be willing to look over any program or even help in the
development of any program for the writing of mathematics in
"Nine-Dot Braille" or any other technical kind of materials,
Mr, Ingham; There couldn't be a better man to do it I
Chairman Rodgers ; Okay. Dr. Nolan has gone and Carl T. Rodgers
is not a research specialist! ,
Mr, Ingham ; Bowing out?
Chairman Rodgers ; Neither a research specialist nor a financier.
Dr, Scott ; In view of the objectives of my own study, I've made
it a policy not to become actively involved in any research programs
in the field other than those involved in sociology and I, therefore,
would not be able to become involved actively in the committee, al-
though I would be interested in monitering its activities and at
the end of three years when I finish my present project, I would be
happy to reconsider it. But at the present time I can't accept
this because of the objectives of my own study.
Chairman Rodgers : I can assure you that after three years have
elapsed there will be plenty of work left to do if you want to
Okay. I think we've covered it all now. I'm very, very
happy to have the specific ways in which this group of participants
would be able and willing to contribute, because this will con-
stitute a tangible set of recommendations from which we can proceed
to try and solve the problems, or at least to look for the people
that can solve them. Also, if any of you now or at any time can
offer any suggestions with respect to funding outside the Foundation,
if you find any money that is available, we would entertain those
I think it has been very fruitful and I was so glad to hear
the comment — I forget who made it — that, after all, this is research
and that there is no reason why we should limit ourselves to any
set of signs or anything else. No one is going to impose anything
on anybody. We're just trying to learn, find out and then, when
the appropriate time comes that we know what we're talking about,
it will be enough time to include the general field and in particular
those who are responsible for the development and standardization
of braille codes, because that's very important, too — we have to
have a stable system — the development and standardization of braille
codes, literary and special,
Mr. Ingham : Before you go away, though, what happens next? We
have all indicated interest and willingness to participate. When
do we hear from you next or what or how?
Chairman Rodgers ; I understand that within two weeks or so we will
get the transcript of this meeting. From it I shall draw up the
Proceedings which will be available for distribution. At the same
time I think we have enough concrete material to seek the interest
of those who would be able to help us, one of them of course being
the Foundation's administration and its Board of Trustees, No doubt
we will get suggestions from the Administration as to other possible
sources of funding, and so we will keep each other in thorough com-
munication, step by step, so that when the necessary funds are avail-
able the various phases of research which you have suggested might
be set in motion. Hopefully, each individual, in consultation with
everybody else concerned, will take on the phase of research in
which he has expressed interest in carrying out.
Dr. Foulke : I would like to add a bit to ray statement. I said
I would be interested in developing an "ideal" code, and I would
consider as a necessary step of doing this, a studv of legibility.
Mr. Ingham : I suppose I'm the only one who hasn't expressed an
Chairman Rodgers : Did I pass you by?
Mr. Ingham : No. I just didn't state what I would be interested
in. Let me just say that I would be interested in approaching our
group leaders at MIT on the whole project, and for myself I would
like to take the development of new codes and testing, especially
the gross testing and distribution and determination of results,
as well as any instrumentation problems which might arise, because
there are lots of facilities available at MIT and elsewhere and
we should be able to tape them. I would be interested in that.
Mr. Covici : I have one question. Should we set some sort of date
by which we have some material, because otherwise things might go
on for a long time. At least some sort of date.
Mr. Strom : A time for which we have--
Mr. Covici : The material to distribute, to send out to people
to have them try to read it.
Mr. Ingham : I think before we Can do that, Bob and whoever else
is going to work on the optimization of this present code have to
Mr. Covici: True.
Mr, Ingham ; That should be the first thing. The duplication
problem worked out and the duplications made, and at the same
time a list of "x" number of people should be available to whom
this material could be sent, graciously received and looked over,
and the questionnaire drawn up.
Chairman Rodgers ; I think Dr. Benham said he could help out
on people available for that purpose.
Mr. Ingham ; But he couldn't be the only one, I suspect, because
much of what Dr. Benham has with his tape library probably consists
of people who are not regular braille readers, of course. Many
of them would be of course, but not all of them. His list should
Mr. Covici ; We could do this immediately,
Mr. Ingham ; Surely.
CONSENSUS OF THE PARTICIPANTS ON THE QUESTION OF PUBLICITY .
The following is a summary of the participants' lengthy dis-
cussion on this subject. With one exception, the consensus was
that publicity should not be avoided altogether, but should defin-
itely be limited to a concise, general announcement, in highly
professional journals, of the June 19th, 196^, "Nine-Dot Braille"
Conference and its recommendations to structure and carry out a
research project on six and "Nine-Dot Braille",
Dr. Nemeth's reasons for opposing any kind of publicity at
this time are as follows:
1. Potential volunteer transcribers will defer becoming
braillists until the outcome of the proposed research
2. Publicity at this time will engender much opposition
from the field of work for the blind.
3. At the present time there is really nothing that can
be reported to any individual or group. When prelim-
inary results are available a summary statement can be
prepared for distribution to interested people.
Chairman Rodp^ers : Thank you very, very much gentlemen, for what
I think are most valuable contributions from each of you.
We are now adjourned.
August, 1965 -88-
"NINE-DOT BRAILLE" CCNFERENCE
JUNE 19, 1964
CARL T. RODGERS
PROGRAM SPECIALIST IN BRAILLE
OTHER TACTUAL EDUCATIONAL MATERIALS
AMERICAN FOUNDATION FOR THE BLIND
" NINE-DOT BRAILLE" CONFERENCE
Friday, June 19, 196^ Helen Keller Room
at 9:30 A»M. American Foundation for the Blind
9:30-9:50 A.M. Welcoming and Introductory Remarks
M, Robert Barnett, Executive Director
American Foundation for the Blind
9:50-10:15 Resume of Efforts in Developing
Robert Strom, Protagonist
10:15 Ten minute break
10:25-12:30 Discussion : Readability of Nine-Dot
Cell characters — Type Scale; Tactually
Discriminable Space Difference of
Modifying Dots (See Items 1-2, Summary
Statement of Working Paper).
12:30-1:30 LUNCH-at the Foundation
1:30-2:30 P.M. Discussion : Character Configuration;
Extension and Position in the Nine-
Dot Cell (See Items 3-6, Summary
2:30-3:00 Discussion : Space-Saving Efforts
(See Item 7» Summary Statement)
3:00 Ten minute break
3:10-^:30 Concensus of the Participants on
the structuring of an effective
research project to determine whether
or not nine-dot cell characters are
feasible from the standpoint of:
b) use for writing by blind individuals
c) small-and large-quantity duplication.
Carl T. Rodgers
"NINE-DOT BRAILLE" CONFERENCE
At the July I96I Convention of the "American Association of
Workers for the Blind", in a paper entitled "Some Developments in
the Various Braille Codes— Literary , Mathematics, Music", I happened
to raise this question: "In light of the continuing birth of new
sciences, new scientific discoveries, new methodologies, and, there-
fore, new terminologies and new forms of established older ones, may
not the time come when the braille system will begin to burst at
its orthographic and legibility seams?"
By taking the number of single dots and dot arrangements (64 —
two to the sixth power) which the six-dot cell permits as the con-
stant factor, and the increasing need to represent new terminologies
and new orthographic styles as the variable factor, we may then con-
ceive of the whole braille problem as one of trying to crack the
six-dot barrier. I believe that Mr. Robert Strom's "Nine-Dot Braille"
proposals represent the first major current effort toward this end.
Only the most extensive and carefully-structured research efforts
will determine whether or not a nine-dot cell is tactually and tech-
Preliminary analysis of Mr. Strom's proposals seems to show a
very close similarity between "Nine-Dot Braille" problems and those
of regular braille. In essense these are problems relating to read-
ability of embossed characters as a function of:
A, Number of dots in a character.
B, Vertical and horizontal extension of characters,
C. Character shape.
D, Characters which are the same in shape, but which
derive their meaning from rotational position changes
in the cell.
E. Characters which are the same in shape but which
derive their meaning from their upper or lower
position in the cell.
F. Type scale (spacing values)
Q. Space-saving efforts through the use of contractions
and short-form words.
A. Number of dots in a character .
The blind participants who evaluated Mr. Strom's original
illustrative material at the first "Nine-Dot Braille" Conference said
that "jumbled" or over-dotted characters cause confusion in discrim-
ination and recognition. While Burklen ( Touch Reading for the Blind )
states that optimal tangibility does not depend on the number of dots
in a character as much as on such characteristics as simplicity of
geometric form, and also on what he called "open characters"
(like the braille letters "m" , "u", or "x"), the tests conducted
by the Uniform Type Committee of the AAWB indicated a higher degree
of legibility with respect to characters containing fewer dots,
from the standpoint of both speed and accuracy ( Proceedings of
the AAWB Conventions— 1907. 1909, 1911« 1913, 1913 ) » The Committee's
tests were conducted with American braille and New York Point, the
two punctographic systems then in use in the United States. The
relationship of character configuration and number of dots in a
character to readability awaits scientific investigation on a
much more thorough level than has hitherto taken place,
B, Vertical and horizontal extension of characters .
In its 1907 report to the AAWB Convention of that summer,
the Uniform Type Committee described a test designed to find out
whether characters two points high were easier to read than chara-
cters three points high. The following brief except from the
Committee's report will enable the group at the Conference to judge
the value of the test: "In fifty-five trials, the list which in-
cluded the tall letters was read in one per cent less time, with
two per cent less errors, than the one in which only the short
letters were used". Burklen's mention of "the easier perception
of open characters" (characters which are always three points high)
may add some support to the Committee's finding on the higher
degree of readability of the "tall" letters.
Because a "Nine-Dot Braille" cell would constitute a hori-
zontal extension of the Louis Braille cell, interest may be added
to the "Nine-Dot Braille" discussion by citing an excerpt from the
1961 Proceedings of the Conference on Research Needs in Braille ,
relating to the Uniform Type Committee's investigation of vertical
versus horizontal signs:
"At the 1909 AAWB Convention the Committee reported on an
experiment designed to find out which signs are more legible — the
horizontal or the vertical ones. Two lists of two hundred signs
each were prepared. In one of the lists all the signs, varying
from one to three dots long, were placed horizontally, while in the
other list the same signs were placed vertically. Each of the
twelve readers who took part in the experiment, all of whom knew
both New York Point and American Braille, read the two lists aloud,
while a seeing person followed the reading with an ink copy,
marking the errors and checking the time spent on each list. The
signs were called not by their letter values buy by the number of
dots they contained. The signs used represented more than one-
third of the recurrence of the alphabet in New York Point and
American Braille, The longest of them could be taken as the ex-
tremes of the two positions — the most horizontal and the most
"The horizontal signs were read in 33 P«r cent more time,
with 321 per cent more errors than the vertical signs. Subsequent
tests gave similar results".
C. Character shape .
See Paragraph A.
D. Characters which are the same in shape but which derive their
meaning from rotational position changes in the cell .
Burklen indicated that certain symmetrical characters of
braille (some 26) differ only in angular position from one another,
causing reversal reading errors, as in the case of the characters
composed of dots 1-6, 1-3, and ^+-3. Examination of Mr. Strom's
revised illustrative "Nine-Dot Braille" material reveals that these
"mirror characters" may occur with equal or greater frequency in
the nine-dot codes than in regular braille, as in the case of the sign
for the word "here", dots 8-1-2-5, which is the regular Braille "r"
rotated 90 degrees counterclockwise.
(The nine dots of the cell are numbered: 7*8, 9i downward on the left;
1,2,3, downward in the middle; and 4,5,6, downward on the right).
E. Characters which are the same in shape but which derive their
meaning from their upper or lower position in the cell .
The Uniform Type Committee designated these as "equivocal
characters". It summarized its findings as follows:
"Experiments tend to show that time is consumed and certainty
diminished by the labor of determining the position of characters
which are like other characters except for their level in a line."
Ashcroft ("Errors in Oral Reading of Braille at the Elementary
Grade Levels") found lower-cell contractions to be fourth (within
a total rank of seven categories) in the order of difficulty from
easy to hard, and stated that "lower-cell contractions were the
most difficult of the group of single-cell contractions."
There seems to be nothing in the literature in contradiction
to these findings. This is worth pointing out in view of the fact
that lower-cell characters occur in "Nine-Dot Braille" at least with
equal frequency as in regular braille,
F. Type scale or spacing values .
This is perhaps the most difficult problem of "Nine-Dot Braille"
For the original material (see paragraph A) the spacing values were:
vertical and lateral distance between two dots
in a single cell O.O8O"
total lateral-cell space O.I6O"
distance between corresponding dots of two
adjoining cells 0.288"
unknown value between adjoining lines of writing.
Some of the participants at the first Conference thought that
the impression of "jumbled characters" resulted from the over-reduced
spacing values between dots in a single cell. The judgement of the
participants may or may not have been correct, but it is interesting
to note that the total lateral cell spacing used for the material
was O.16O", or O.O7O" greater than the lateral cell spacing (O.O9O")
of a regular braille cell, and the distance between corresponding
dots of two adjoining cells was O.O38" greater than that of two
adjoining regular braille cells, which is 0.250".
Each of the three "Nine-Dot Braille" slates which were con-
structed for embossing new material represents a different spacing
value as follows:
First slate. Distance between dots in a single cell. . . .O.O85"
Second slate. Distance between dots in a single cell. . . .O.O9O"
Third slate. i^istance between dots in a single cell. .. .0.095"
The spacing between cells and between lines was the same for
the three slates; namely, O.I6O" between cells and point 0.220"
As indicated in the "Addendum Sheet", the addition of the
modifying dots 7, 8, 9» at the left of the cell is unfeasible
without an increase of "x" decimal of an inch between the modifying
dots and dots 1,2,3, the mid-cell dots. For instance, the "Nine-
Dot Braille" contraction for "ast", dots 3t^-7, can be read as
the regular braille "ch" sign (dots 1-6) followed by dot 4; the
contraction for ""full", dots 1-2-4-7-8-9, can be read as the
regular braille "q" followed by dot 4; the contraction for "ound",
dots l-A--5-7-9» (the regular Grade Two contraction written as a
one-cell contraction), can be read as the regular braille "m"
followed by dots 4, '5, and so on.
The reason for these ambiguities is obvious* The distance
between dots 3-7 of the "ast" contraction equals the distanice
between dots 1-6 — the regular braille "ch" sign; the distance
between dots 1-2-7-8-9 of the "full" contraction equals the dis-
tance between dots 1-2-5-^-5 — the regular braille "q"; the dis-
tance between dots 1-7-9 of the "ound" contraction equals the dis-
tance between dots 1-3-4 — the regular braille "m".
Many more examples of ambiguities and distortion of regular
braille configurations can be cited, but it is' believed that enough
has been said to show the need for an increase of "x" decimal of
an inch between the modifying dots and the remainder of the cell.
Since this appears to be true, and since the spacing values which
were used in the original "Nine-Dot Braille" material seem to
have disturbed a number of readers at the first Conference, the
following figures on spacing variables are offered in the hope of
stimulating more thinking on the problem.
Regular Braille Cell
Distance between corresponding dots of two
adjoining cells 0.250";
between corresponding dots of adjoining lines,. 0.400".
Number of cells in a 10" long line 40.
Number of lines on 10" high writing space 25.
Total number of cells per page,...,,,... 1,000,
(Until otherwise indicated, the distance between two lines
of writing should be taken to be 0.220".)
Distance between corresponding dots of two adjoining cells
when the distance between dots in a single cell is O.O8O",
and the distance between cells is 0.123": O.283".
Number of cells in a 10" long line 35.
Number of lines in 10" high writing space 25.
Total number of cells per page 875.
Distance between corresponding dots of two adjoining cells
when the distance between dots in a single cell is 0.090",
and the distance between cells is 0.123": 0.303".
Number of cells in a 10" long line 33.
Number of lines in 10" high writing space 25.
Total number of cells per page 825.
Distance between corresponding dots of two adjoining cells
when the distance between dots in a single cell is O.O8O",
and the distance between two cells is 0,l60": 0.320".
Number of cells in a 10" long line 31.
Number of lines in 10" high writing space 25.
Total number of cells per page 775.
Distance between corresponding dots of two adjoining cells
when the distance between dots in a single cell is O.O9O",
and the distance between two cells is O.I6O": 0.340".
Number of cells in a 10" long line. 29.
Number of lines in 10" high writing space 25.
Total number of cells per page 725.
When the preceding four lateral variables are applied in the
same order, but the vertical distance between dots is reduced to
0,072", and the distance between the lines is reduced from 0.220"
to 0.163", then the number of lines on a 10" high writing space in-
creases from 25 to 32, and the cell totals per page are as follows:
1,120; 1,056; 992; 96O.
The figure 1,000, which is the total number of cells per
page for regular braille using a 10 x 10" writing surface, might
serve as the constant factor when studying the readability of
"Nine-Dot Braille" as a function of "x" spacing variables — lateral
and vertical. In offering these figures the dot-base diameter has
been assumed to be 0.053". Changes in dot diameter, and possibly
also in dot height, might help increase the legibility of nine-
dot cell characters.
The foregoing figures were suggestea by the 1920 Report
of the AAIB-AAWB Commission on Uniform Type for the Blind, and by
the study on spacing variables of Meyers, Ethington and Ashcroft
("Readability of Braille as a Function of Three Spacing Variables").
Attempts to ascertain the spacing values for embossing New York
Point — which in some respects resembles the nine-dot cell — have
thus far been futile.
Two more considerations with respect to spacing problems
seem appropriate. The first relates to diagonal dot distances.
Since both the six-and nine-dot cells are quadrangular, these
distances may be calculated through the formula "a" square plus
"b" square equals "c" square. The diagonal distances yielded are
Distance between dots 1,6 or dots 3»^ 0,201".
Distance between dots 7, 6 or dots '+, 9 when the distance
between both vertical and horizontal dots is O.O9O" . . . .0.25^"
Distance between dots 7, 6 or dots 4-, 9, if the lateral
distance between dots is reduced to O.O8O" and the
vertical distance is reduced to 0.072" 0.215"
The second consideration is concerned with the relationship
of dots 7, 8, 9, of the nine-dot cell to dots ^, 5, 6 as modifying
dots in regular braille two-cell signs. In regular braille, words
and letter-groups are contracted into a two-cell sign by prefixing
one or several dots to a letter or a contraction — dots 5, 1-5
for "ever", dots 4-5-6,2-3-4-6 for "their" and so on.
Query: Aren't dots ^, 5i 6 in effect functionally identical
with dots 7, S, 9i in a "nine-dot cell", except for a larger space
between them and the remainder of the signs of which they are a
If investigation reveals that "Nine-Dot Braille" type, in
order to be tactually feasible, must exceed the 0.1 square-inch
surface scale of regular braille, is the solution to the problem
to be found in a more extensive use of contractions?
G, Space-saving efforts through the use of contractions and short -
form words .
This has been one of the most highly controversial aspects
of touch reading. Limited investigations by the Uniform Type
Committee, the Commission on Uniform Type for the Blind, and Ash-
croft, reveal that certain contraction categories tend to facilitate
reading, while others tend to produce various types of reading
errors. There is no reason to believe that in this respect, except
perhaps for the possibility of substituting contractions for short-
form words, "Nine-Dot Braille" contraction problems would be any
different from those of regular braille. Only through additional
and extensive scientific investigation can one hope to arrive at
suitable solutions to the whole touch-reading problems. One would
also hope that in the continuing development of embossed systems
for blind persons, this principle will always be rigidly follows:
It is the needs of the blind reader — not of the system that must
be met .
The enormous scope of the subject in itself has precluded the
possibility of a truly suitable working paper. For one thing, only
a cursory review of the investigation records available was feasible,
since any well-structured research project would require a far
lengthier review of the literature than could be presented here.
The purpose of the foregoing material has been simply to attempt
to set up the kind of questioning frame of mind that it is hoped
fully will lead to a fruitful, analytical discussion at the on-
coming "Nine-Dot Braille" Conference.
SUMMARY .STATEMENT OF SUGGESTED RESEARCH
1. Readability of nine-dot cell characters as a function of type
scale — dot size, dot height, lateral and vertical distances
between dots in a cell, distance between adjoining cells, dis-
tance between adjoining lines of writing,
2. Readability of nine-dot cell characters as a function of
tactually discriminable space difference between dots ?♦ 8, 9i
used as modifying dots at the extreme left or right side of
the cell, and the remainder of the cell.
3. Readability of nine-dot-cell characters as a function of
character configuration and character extension--lateral and
4. Readability of nine-dot-cell characters as a function of
characters which are the same in shape, but which derive their
meaning from their lateral position in the cell — left, middle,
right -nosition; left-middle, middle-right position,
5. Readability of nine-dot-cell characters as a function of chara-
cters which are the same in shape but which derive their mean-
ing from their upper or lower position in the cell,
6. Readability of nine-dot-cell characters as a function of
characters which are the same in shape but which derive their
meaning by being rotated as much as 3^0 degrees.
7. Readability of nine-dot-cell characters as a function of greater
space-saving efforts than those prescribed in the "AAIB-AAWB
Braille Authority's English Braille-American Edition 1939-
« * 1^ iK « 4i
"NINE-DOT BRAILLE" CONFERENCE
AMERICAN FOUNDATION FOR THE BLIND
JULY 6, 1962
Carl T. Rodgers
Program Specialist in Braille and Other Tactual Aids
Miss Kathern F. Gruber, Chairman
Director of Division of Program De-
American Foundation for the Blind
Mr, Kenneth Ingham
Mathematician-Graduate Student at
Mr, Howard M, Liechty, Editor of
the "New Outlook for the Blind" and
Member of the Board of Trustees of
the American Foundation for the Blind
Mr, Carl T. Rodgers
Program Specialist in Braille and
Other Tactual Aids
American Foundation for the Blind
Mr, George F. Meyer, Executive Director Dr, James Slagle
New Jersey State Commission for the Blind Research Associate
and Member of the Board of Trustees of Massachusetts Institute of
the American Foundation for the Blind Technology
Dr, Thomas A. Benham
Professor of Physics
Haver ford, Pennsylvania
Mr, Robert Strom
Inventor and Protagonist of the
"Nine-Dot Braille" and an
undergraduate student at
Mr. Robert A, Bowers
American Foundation for the Blind
Mr, Arthur L. Voorhees
Program Specialist in Vocational
and Rehabilitation Services
American Foundation for the Blind
Mr. John K, Dupress, Director of
American Foundation for the Blind and
Research Associate of the Department
of Electrical Engineering
Massachusetts Institute of Technology
Dr. Everett E, Wilcox
Program Specialist in Education
American Foundation for the Blind
Miss Joan M, Stapleton, Secretary
American Foundation for the Blind
PARTICIPANTS INVITED — UNABLE TO ATTEND
Mr, Ted Glazer
Manager, Systems Division
Burroughs Research Center
Mr. M, Robert Barnett
American Foundation for the Blind
Mr, Louis H. Rives, Chief
Division of Services for the Blind
Office of Vocational Rehabilitation
Dept. of Health, Education, and
Washington 25, D.C.
"NINE-DCT B'^AILLE" CCNF^iRENCE
This Gonferenc* was held on July 6, 1962, from 10:00 a.m. to
^:00 p.m. in the Helen Keller Room of the American Foundation for
the Blind. Miss Gruber opened the Conference by stating its
purpose as follows:
One of the major responsibilities of the American Foundation
for the Blind is to evaluate all ideas, systems, methods,
aids, etc., that are formally ^resented to the Foundation
for this purrDose. Within the past two months, Mr. .Robert
Strom -presented to the Foundation a complete system of
"Nine-Dot Braille" for evaluation. The Program Sn^^cialirt
in Braille, Mr. Carl T. Rodgers, made the pr^^liminary in-
vestigations, and as a result of his findings, he recommended
to the Foundation administration that a small group of tota''lv
blind re=5ders be convened for the pur-nose of further analysis
and ex-nloration of this "Nine-Dot Braille.
Miss Gruber then established a most satisfactorv auditory
rap-port araonfr those present by having each of them make a brief
statement of self-identification. Following this, the Program
Specialist in Braille described the content and the arrangement
of the Conference material; told how it had been prepared and d\-
plicated: explained some of the technical difficulties which were
encountered in producing the material on a somewhat crude prototype
"Nine-Dot Braille" slate; and stressed that all of the samples and
the spacing values used were tentative and subject to whatever
chanp-es and revisions mip-ht be indicated by reader reaction and
adequate research studies.
GENERAL DIS CUSS ION TOPICS :
1 . Legibility and Spacing Values of "Nine-Dot" Characters
It was observed by the participants that "bunching of dots" and
widely-spaced, isolated dots decrease reading facility. Among such
signs were the "anv" sign (dots 1-2-3-^-6-7-8-9), the "ity" sign
(dots 1-3-^-^-6-7-8-9), and the "an" sign (dots 1-9).
It was also noted that "Nine-Dot Braille" characters which did
not "overcrowd" the sense of touch a>"e more easily read. Among
such symbols were the "is" sign (dots 2-^-9), the "ate" sign
(dots i-8-9), and the "ly" sign (dots 1-2-3-7-9).
It was suggested that some of the signs which appeared "bunched
together" might be made more legible by appropriate alterations
of the spacing values of the nine-dot cell, and it was agreed that
o-otimal spacing values would have to be established through ad-
equate research. It was stated that the findings of the study en-
titled "Readability of Braille as a Function of Three Spacing
Variables" by Meyers, Ethington and Ashcroft was a fine contribution
to the field of research in braille and a good starting point for
further investigations concerning suitable spacing values, dot
diameter, and dot height.
2. The Writing of "Nine-Dot Braille "
Mr. Strom's "Nine-Dot Braille" slate was examined by the par-
ticipants. It was pointed out that dot 5i being the middle point
of the square, (the nine-dot cell) might be difficult to write,
to which Mr. Strom, who had punched out all of the "Nine-Dot Braille"
Conference material replied that he had encountered no difficulty
with the writing of characters containing dot 5. Brief mention
was also made of a keyboard suitable for writing "Nine-Dot Braille"
as well as for manual operational requirements.
3. " Nine-Dot Braille" Structure
The Nine-Dot cell was developed simply by adding the column of
dots 7,8,9 to the right of the conventional braille cell. It was
suggested at the Conference that the added column might be placed
at the left, rather than at the right of the conventional cell for
the purpose of using combinations of the dots of the added column
much the same as dots ^, 5« 6 of the standard cell are presently
used in compound symbols. This would retain, more compactly, the
already existing symbols of Grade 2 braille, avoiding obsolescence
and affording the possibility for the creation of new symbols. Mr.
Strom and the Program Specialist in braille stated that this al-
ternative for develo^ ing nine-dot characters might be quite feasible
since the nine-dot characters would then be truly an extension of
the six-dot cell characters. Mention was made by Mr. Meyer of the
possibility of a variable cell in order that some of the characters
might occupy less space than the full nine-dot cell. It was felt
by some of the participants that this idea, which would be feasible
through a properly constructed slate or machine, might serve to
save considerable space and lend continuity to the legibility of
^. Purposes and Needs of Braille
Mr. Meyer, addressing himself directly to the Program Specialist
in braille, asked him to state his "impressions" concerning braille
needs. The following summarizes Mr. Rodgers ' effort to answer Mr.
"First, I believe that as a system of touch reading and
writing for blind persons, braille is a multi-purpose system.
When we speak of 'needs', what needs are we talking about? Are
we talking about the needs of the reader who wants to read for
recreation only? Are we talking about the needs of the *John
Covicis' who need an orthographic system for the accurate re-
cording of current mathematical and scientific concepts and data?
Consider the orthographic discrepancies that exist between the
symbols provided for in Braille Chemical Notations and How to
Use Them by Loomis and Mitchell, and the symbols of the Nemeth
Code. At the present time, braille mathematical and scientific
notation is in a state of complete fuidity. Are we talking about
the needs of the blind musician, whose present braille notation
is highly technical and multi-celled? When speaking of 'needs'
we have to break them down the way we do when dealing with the
needs of any handicapped group -- into highly individualized
needs. Each available medium — braille, talking book records,
tapes — must be used to meet the needs of individuals . "
The rest of the discussion on the question of needs revealed
the feeling that a nine-dot system would be welcomed by "the
blind individual seeking a way in which to record highly special-
ized work", but that "the, individual who reads non-technical
braille is going to rebel,"
SOME CRITERION TO U3E IN DEVELOPING NINE-DOT CHARACTERS
The following is a summary of the ideas offered by the par-
ticipants to give order and meaning to the potential creation of
1, Wherever possible there should be strict adherence to
the principle of orthographic representation (each symbol
to be assigned one meaning).
2, Concentrate on developing characters not now available
in braille for representing certain existing ink-print
3, While absolute standardization of symbols and symbol
meanings does not really exist in inkprint, it would
seem wise to establish code uniformity by providing
braille equivalents for those inkprint symbols used
most commonly and having the least variations of meaning.
h. Whenever possible use the right side of the "Nine-
Dot Cell" for the special symbols of mathematical
and scientific notation, using the left side of the
cell only when it becomes necessary to avoid con-
fusion with a symbol of the literary code.
5. New characters should be created according to a
consistent pattern or system of development,
6. "Nine-Dot Braille" characters should be developed in
keeping with logical inkprint presentation. For
example the representation of the inkprint notation
which means "a-sub-one" should be designed so as to
read exactly like that, and not "a-one-sub."
7. In developing "Nine-Dot-Braille" make it as compatible
as possible with the old system so as to reduce the
necessity of learning an entirely new system.
8. The establishment, through scientific experimentation,
of optimal spacing values as an indispensable re-
quisite to the legibility of nine-dot characters.
9. Experimentation with other types of cells, such as
2 points vertical and k points horizontal, etc.,
should be included in future research on touch reading,
BRAILLE 3YMB0LS FOR "PRINCIPIA MATHEMATIC" AND MATHEMATICAL LOGIC
For the purpose of illustrating the urgency for creating a system
of touch reading and writing that can adequately meet the require-
ments of modern symbology, Mr, Strom read a passage from the book
Principia Mathematica, a reading "must" for students of higher
mathematics, and which cannot be made available to blind mathemati-
cians because the existing mathematical braille notation provides
no orthographic equivalents for the symbols used in this book. Mr.
Strom stated that suitable single-cell equivalent svmbol":^ could
be provided by "Nine-Dot Braille", for not only Principia Mathematica
but mathematical logic as well. "
The material prepared by the Program Specialist in Braille in-
cluded the four following questions, the answers to which sum up the
participants' consensus of opinion on what action, if any, the
American Foundation for the Blind might take regarding Mr. Strom's
"Nine-Dot Braille" proposal:
1. In general, does it appear that the reading finger can s-oan
and comprehend Nine-Dot characters?
Dr. Benham Yes,
Dr. Slagle Yes.
Mr, Dupress Yes,
Mr. Ingham Yes, but samples of "Nine-Dot Braille" should
be produced with more adequate writing equip-
ment in order to be able to judge better,
Mr. Meyer Yes, it could be.
2. Does it appear that some of the characters are especially diffi-
cult to read?
Dr. Benham Yes, some are hard to rend, but if dot height
and spacing values were uniform, they would be
easier to read. Uniformity and better character
designing (placing indicators before instead of
after the basic symbols) would, I think, make
the system easy to read,
Dr, Slagle Yes, characters that have many dots are diffi-
cult to read. The use of dot 5 might, in
general, be avoided when you have a choice.
Sparsely-located dots cause difficulty,
Mr. Ingham Yes, characters with too many dots are hard to
read. Also, the column of dots 7*8, 9i might
make for easier readinp if added at tne left
instead of at the right of the basic cell,
Mr, Meyer Yes, difficulties increase when you consider
adjacent characters. Because of the spacing
values which have been used for this material,
you can't tel] where one cell leaves off and
the other begins.
Mr. Dunress — I will agree with the group that where there
are only a few dots, left or right, the charac-
ters are difficult to read. There is need for
research on this,
3. Would it be desiral.le to treat the "Nine-Dot Cell" as an ex-
tension of the "Six-Dot Cell" in developing nine-dot symbols?
All a.p-reed in the affirmative, since this would avoid render-
ing the old system obsolete, would eliminate the need for re-
learning many new characters and would, therefore, result in
potential subjects being more willing to cooperate in field-
^, From the standpoint of present-day orthographic needs on the
one hand, and of tactual legibility requirements on the other,
does the "Nine-Dot Braille" concept appear sufficiently pro-
mising to justify undertaking the extensive and comprehensive
experimentation, field-testing, and other long-range steps
necessary for the development of the proposed concept into an
optimal, multi-purpose system of touch reading and writing?
All agreed that additional investigation must precede any large-
scale experimentation and other long-range action. As part of
further -oreliminary investigation, the participants recommended
a better-constructed, more sophisticated "Nine-Dot Bra-'"'!*"
slate to afford better evaluation of the readability of nine-
dot characters. A few Foundation people and a few well
selected outsiders should continue the investigation. Further-
more, a second conference on "Nine-Dot Braille" was recommended,
at which time a definite decision could be made in answer to
The protagonist of "Nine-Dot Braille" posed the ouery: "Do
you think v/e can recommend experimentation as to the size of
the cell and s];)ecific values?" The unanimous reply was
Finally, everyone recommended gradual and thorough investi-
gation, evaluation and data gathering of the proposed system, with
emphasis on technical needs for the benefit of the blind student
of higher mathematics in particular and technical training in
general. Stress was laid on making no announcements or taking any
action which might lead to a renewal of a "War of the Dots" or
"Battle of the T;/-Des" among the users of braille or among the
workers in the field.
MATHEMATICS GF SIX-DOT BRAILLE
AND EXPANDED PUNCTOGRAFHIC CODES
FOR TOUCH READING AND WRITING
Submitted by: Dr. Abraham Nemeth
Assistant Professor of
THE MATHEFATICS OF SIX-DOT BRAILLE AND
EXPANDED FUNCTOGRAFHIC CODES FOR TOUCH
READING AND 'SITING
Submitted by: Dr. Abraham Neraeth
Assistant Professor of Mathematics
A. COMPARISON OF SPACE REQUIREMENT 3 OF SIX-
DOT BRAILLE WITH "NINE-DOT BRAILLE".
Betv/een dot centers within a single cell OQO inches
Between right side of one cell and left side
of next cell l60 inches
Between bottom of cell on one line and top of cell
on next line 220 inches
In normal six-dot braille, it is possible to emboss 40 cells
to the line and 25 lines to the p-^.ge if the dimensions of the p ige
are 11 inches by 11>^ "'nches while still maintaining suitable margins.
With the same spacing parameters for "Nine-Dot Braille", it is poss-
ible to emboss 29 cells per line and 25 lines per page while still
maintaining suitable margins. There :-.re thus 1,000 cells per page
available in six-dot braille compp.red with 725 cells per page avail-
able in "Nine-Dot Braille", Proper comparison of the two systems
requires that the spacing parameters be assumed equal in both.
It is assumed that ordinary adult prose has been transcribed
in the two systems being compared, .Because of the variable length
of words and syllables, it is not possible to emboss either all of
the 1,000 cells per page in six-dot braille or all of the 725 cells
per page of "Nine-Dot Braille", It is estim.^ted that an average of
1.6 cells per line are unused at the ends of lines for these reasons.
Accordingly, there will be 40 cells per page lost in either svstera.
There will thus be 960 cells per page available for six-dot braille
and 685 cells per "page available for "Nine-Dot Braille", The 40-
cell-per-page looS constitutes a 4 per cent loss in six-dot braille,
but a 5.5 per cent loss in "Nine-Dot Braille", We will continue to
assume, from this point onward, that there are 96O cells and 685 cells
per page, resDectively , available for braille transcription in the
It has been pro-nosed that a nine-dot cell be regarded as com-
posed of a left third to accommodate either com.position signs or
"prefixes" of two-cell contractions, and a right two-thirds to accom-
modate letters or part-word signs affected by such composition signs.
or the "bodies" of two-cell contractions. Four of the six
composition sip:ns of English Braille can be accommodated in the
left third of a nine-dot cell. These are the capital, italic,
letter, and accent signs. The number sign and the termination
sign cannot be accominodated in this way.
Each two-cell combination of composition sign followed by
a single letter or part-word sign, or of a prefix followed by
the body of six-dot braille can be accommodated in a single cell
of "Nine-Dot Braille", so that there is a gain in "Nine-Dot
Braille" of one efficiency unit. In order for "Nine-Dot Braille"
to be more efficient than six-dot braille with respect to the
amount of s-nace occupied, "Nine-Dot Braille" must score a gain
in excess of 275 efficiency units, since this is the number by
which the cells per page in "Nine-Dot Braille" falls short of
the cells per page in six-dot braille.
In English Braille, Grade Two, there are ^7 two-cell
contractions. A survey of adult prose material suggests that
the frequency of occurrence of such contractions is at the rate
of about 25 per page of six-dot braille. Since there are seven
prefixes and 5^ bodies available, it is theoretically possible to
construct 7 x 56 or 392 two-cell contractions, as in Grade Three.
It will be assumed that the rules for forming contractions can be
formulated in such a way that no conflicts arises from the dual
role played by four of the prefixes as composition signs. Whether
the increased number of contractions acts to slow down or to
speed up the reading;- rate is a psychological matter for investi-
gation, and is entirely disregarded here. Thus, when 3*^2 is
compared with ^7i it is seen that the incidence of two-cell
contractions could be about eight and one-third times as great
as is the case in grade two. It will therefore be assumed tViat
the density of two-cell contractions would be eight and one-third
times as great when using all the possible 392 two-cell contractions
than is the case when using the 47 two-cell contractions of English
It may therefore be expected that the frequency of occurrence
of two-cell contractions would be at the rate of (8 1/3) times 25
or about 209 occurrences per page of six-dot braille. This is a some-
what generous assumption, since it is likely that the first ^7 two-
cell contractions of English Braille have a greater frequency of
occurrence than the remaining possible 3^5 two-cell contractions.
A survey of adult rirose material suggests that the occurrence
of the capital, italic, letter, and accent signs combined is at
the rate of about 15 per page of six-dot braille, grade two. If
the fully contracted system of 592 two-cell contractions were used,
these composition si(2;ns would occur in 751 (that is, 9^0 - 209)
cells instead of in 9^0 cells, as in grade two braille. At the
rate of 1^ occurrences in 751 cells, there would be 19 composition
signs per page of highly-contracted six-dot braille.
The occurrence of the double capital sign or of the double
italic sign might also be taken into account. However, the single
capital sign followed by a two-cell contraction can result in a
gain of only one, and not two, efficiency units in nine-dot braille.
Because these two factors tend partially to cancel each other, and
because the occurrence of either is relatively rare, we neglect both
The 209 occurrences per page of two-cell contractions in highly
contracted six-dot braille, plus the occurrence of 19 composition
signs per page result in a gain of 228 efficiency units when the
same material is embossed in "Nine-Dot Braille". As was indicated
earlier, a gain of 275 efficiency units is required just to break
even. The gain of 228 units of efficiency thus falls short of the
break-even point by ^7 efficiency units. We must therefore conclude
that if the same spacing parameters and the same set of two-cell
contractions are used both in six-dot and in "Nine-Dot Braille",
six-dot braille still has an advantage over "Nine-Dot Braille" with
respect to s-nace requirements.
An alternative suggestion has been made that the "prefix-body"
concept be abandoned, and that each nine-dot character be made to
represent a contractable combination in its own right. From this
point of view no more contractions are available than in the prefix-
body concept, and its application can only result in a permutation
of the available nine-dot characters for the representation of the
desirable contractable combinations.
If each of the spacing parameters is reduced by .010 inches,
it would be possible to emboss 42 cells of six-dot braille per
line compared with 31 cells of "Nine-Dot Braille" per line. It
would also be possible to emboss 2? lines per page in each system.
With the new parameters there would thus be 1,13^ cells per page
of six-dot braille compared with 837 ce]ls per page for "Nine-
Dot Braille". Again allowing a loss of 1.6 cells at the end of
each line due to the variable length of words and syllables in
each system, there would be available 1,091 cells of six-dot
braille per page compared with 79^ cells of "Nine-Dot Braille"
per page and the break-even point would come with a gain of 297 units
of efficiency. The increase from 960 cells to 1,091 cells per page
represents a gain o-^ about 13.7 per cent. If th^re are 128 occur-
rences of two-cell contractions and composition signs usinp- the
standard parameters, then, at the same rate, there would be 25Q
such occurrences using the smaller spacing parameters. "Nine-Dot
Braille" would thus provide 25^ units of efficiency per page but
this is still 38 units of efficiency short of the break-f-ven -^ojnt.
We must again conclude that six-dot braille still rpnain.-^ no--f^
efficient than "Nine-Dot Braille" with resp-ct to s-ace rem i -rprr^nts
even when the smaller parameters are used.
The apparent advantage of "Nine-Dot Braille" over six-dot
braille which is evidenced in the sample materials distributed
to the conference participants comes from the fact that the nine-
dot transcriptions were made using both the smaller parameters as
well as a hij:;hly contracted system, whereas the six-dot braille
with which it was compared used standard parameters together with
the standard contractions of Grade Two,
CCMFARISON OF INTRINSIC AMBIGUITIES IN
SIX-DOT AND "NINE-DOT BPAILLE".
As is well known, many of the char'.cters of six-dot braille are
intrinsically ambiguous, and that their precise orientation within
the cell is often indicated by the placement of a full cell of dots
in an adjacent position. In the reading of standard prose material,
such ambiguity is almost entirely resolved both by the presence of
adjacent characters which serve as orienting symbols, and by the
context of the material itself. What role each of these factors
rlays ir the resolution of these ambiguities is not known. We shall
here be concerned only viith intrinsic ambiguity, and both of the
factors which resolve it will therefore be absent.
Characters vary with respect to their degree of ambiguity. Thus
a one-dot character has six degrees of intrinsic amhie-uity; the
letter b has four degrees of ambiguity, since, without orientation
and context, it could be taken as dots 1-2, 2-3, ^-5i or 5-6; the
letter c has three degrees of ambiguity since, without orientation
and context, it could be taken as dots 1-^, 2-^, or ^-6; the letter
e has two degrees of ambiguity since, without orientation and context,
it could be taken as dots 1-5 or 2-6; the letter m has but one degree
of ambiguity, since it can only be taken as dots 1-3-^. It is thus
clear that a rating of one constitutes no ambiguity at all. It is
proposed, then, to neglect the space, and to rate the ambiguity of
each of the remaining 65 characters of six-dot braille and the remain-
ing 511 characters of "Nine-Dot Braille". The average of these ratings,
obtained by dividing the sura of the ratings nn each svstem by 63 and
by 511, res-oectively , will then be a measure of the intrinsic ambiguity
of each system and these intrinsic ambiguities may then be comp^'red.
We use the letters A, B, and C to designate the dots in the l^-ft
third of a nine-dot cell from ton to bottom, respectively; the numbers
from 1 to 6 are used for the remaining dots of the nine-dot cell in
complete correspondence with the dot numbering of six-dot braille. In
"Nine-Dot Braille" a one-dot character ha- nine degrees of ambiguity;
a character consisting of dots A-B has six degrees of ambiguity since
it could be taken as dots A-B, B-C , 1-2, 2-3, ^-5, or 5-6; etc.
While it is nossible to assign an ambiguity rating to each
character of each system by inspection, it is also possible to develop
a formula for t^-is purnose. With each character there is associated
a certain number of exterior blank columns and exterior blank rows.
To find the number of exterior blank columns associated with anv
character, start from the left edge and from the right edge of the
cell and count the number of blank columns which are r^resent. For
example, the character which consists of dot 1 in six-dot braille
has one blank exterior column; the same character in "Nine-Dot
Braille" has two blank exterior columns. The character consisting-
of dots 1-3-^ has no blank exterior columns in six-dot braille, but
has one blank exterior column in "Nine-Dot Braille". The charac-
ter consisting of dots A-C-4-6 in "Nine-Dot Braille" has no exterior
blank columns althoup-b the middle column is blank. In a similar wav,
to find the number of exti^rior blank rov/s associated with a charac-
ter, start from the top edge and from the bottom edge of the cell,
and count the number of hi ark rows which are present. For example,
the character consisting of dot 1 has two blank exterior rows in
both systems, the ch-^.racter consisting of dots 1-^-^-6 has no blank
exterior rows in either system even though the middle row of this'
character i.s blank.
We let c denote the number of blank exterior columns associated
with a character. In six-dot braille, c can have one of the values
or 1; in "Nine-Dot Braille" it can have one of the values 0, 1,
or 2. Similarly, we let r denote the number of blank exterior rows
associated with a character. In both systems r can have one of the
values 0, 1, or 2. If R denotes the ambiguity rating for any charac-
ter, then the formula for "R is R=(c +l)(r + 1), For example, the
character consisting of dot 1 in six-dot braille has associated with
it one blank column and two blank rov/s. Therefore for this charac-
ter, R = (1 + 1)(2 + 1)=2 X 3 = 6. The same character in "Nine-Dot
Braille" has associated with it t/o blank columns and two blank rows,
so that R = (2 + 1)(2 + 1) = 3 X 3 = 9 for this character in "Nine-
Dot Braille", The formula for R holds equally well for any rect-
angular cell such as the two-by-two cell of New York Point and the
four-by-two cell of the German shorthand system.
The tables which follow list the 63 characters of six-dot braille
and the 311 characters of "Nine-Dot Braille", together with the
ambiguity rating of each character in each system. It is found that
the average ambiguity for six-dot braile is 2.05 and that the corres-
ponding rating for "Nine-Dot Braille" is 1.72. We, therefore, conclude
that "Nine-Dot Braille" has less intrinsic ambiguity than six-dot
braille and that the ratio of the ambiguities is 1.72/2.05, or .84;
"Nine-Dot Braille" is only .84- as ambiguous as six-dot braille.
INTRINSIC AMBIGUITY OF SIX-DOT BI?AILLE
NOTE: In the following table the column headed (1) specifies
the six-dot braille characters by dot numbers; hyphens are omitted
between these numbers. The column headed (2) specifies the ambiguity
ratins- of the corresron ding character in column (l).
(1) (2) (1) (2) (1) (2)
t — '
f\J N^i J- lAVO r<^^ lAMD ^lAV£)mvX>VOnjKN^lAv^hrxJ-irN
VO ITNVOVO (\J KS^ lAVD K\4f lAVD^ ir\»vD lAV£)>sO r^^ lAvO
O r-l (M r<^ ^ IAV£) iH <M K>^ lAVO f\J rr\ ^ ir\*Xi r<~\-d- lAVD J-
0^a^0^a^0^a^O^O^a^VO KNVDJ- f\j hf>(\l HVX)^vi5^ OJ K~\
„ PQ O H OJ KN-d- ITNVT) O H f\J KN J- IA
PQ O H OJ K>^ lAVO <a;<C«a;<i;<a;<!«aJ<ajpQpQpQpqpQpQ
>sD vX) v^ ^£> LTx VT) i^£)
Lr\ ltn ir\ lA J- J- lA
LTn VX) VX) \X) -d- ITNMD
J- ^ lA ir\ K^ r<~\ tox
•sDd" lAO m^sOvO KNcf LTWDd- lA^D lAVDV£l-:f Lr\V£) LTN^X) r<~\rAK^iJ- OJ OJ fV'
(\]K-xfAK\-d-j-Lr\nj(\i(\jf\jK\K\N--, j-j- L^^K^K^K^J-^ooooHr^H
kd [Tskd \d ^ ^ ir\
■ - - lAvO irwO VX5 4" LA
lA PA-d- -4- LA rA N^
H r-l H
v«.«.« ^-^ ^*^ lAVOvX) lAVOVOvD lAVDvOMDVD ITvVOVDVOvDVD lr^VD^^VDvr)
H H H H H H H
r-i t-i r-\ r-i
rH H iH H H
VOIAVDvD lAVO>^VO irNv£)>X)V^vr> 4- lAvO lAvDVD lAvOv£iVO lAVD VX) VD VO [TwD
^■^ •3'. "^^ J-lTMA-d-^-irMTMrvN^KN Kn J- 4- IT, J- 4" ITN ITnJ- J- lA IT, lf\ J- J-
iH H H iH iH H
(H H H H iH
K>cr4- ir\4--d- irsir\rr\f<>K\^j- ir\^^ iAir\4-j- irstr\ir\ooooc_5oo
vi3vX)ir\v£)vDv£)V£iv£)v£)^-d--:l- LT, IT, Lr\Lr\miALrN^
LTx LP <f ^ LA IT > iTN L^^ LP- ir\ r^ K~', r^ -d- -cf -d- J- -d- -d- f^
J- J- K> K, K- J- -4- -c^ -d- -4- (M fv oo rv r<~x r^ K> K\ K^ fM
H H iH H iH
OJ rj fV' Tv: (\J H H H (H H rH f— I H rH H I— I I— I rH H
VD vo VD vx. f\J
LT LT LT LT
J- LAVDJ- LP, V£) U^,V£)V£)-:t- LAVX) ITWOVX)
OJ f\i f\.
fi- t^ t^
LT^ K\ K\ r<^ ^ -:*■ LA
rH fVi (\J f\J OJ f\J f\J
j-cj- ir\iAmr\j cm c\j (\j r^rAi^. -:t -d- lakxknknj- ^
H H H
H r-H rH rH
OJ ru OJ (H
LPWT)^ invX) lAVDvO K\^ lPvVX)J- U^VD IAvX)VC-:l- IPWD
CONFESLKCS ON "NINE-DOT BHAUXE",
Proceedings, N'i'C AF:B, June 19,
196h. ' '